Zscaler, Inc. Aktie

Welcome to Zscaler Investor Day. Please welcome to the stage Zscaler SVP of Investor Relations and Strategic Finance, Kim Watkins.

Thank you. Hello, everyone. It's my pleasure to welcome you to Zscaler's 2026 Investor Day. And whether you're here with us in New York or you're watching on the webcast, we really appreciate you taking the time to be with us today.

So it has been 5 years since we hosted an Investor Day. During that time, we've grown into a broad and integrated platform built on our Zero Trust Exchange. We have a great lineup today to tell you more about our evolution and to help you come away with a better understanding of our strategy, our product differentiation and our growth drivers.

So Jay will kick us off and take us through Zscaler's vision and strategy, including how our solutions fit into the changing cybersecurity landscape. Next, Adam will take us back to basics and introduce the Zscaler platform, followed by Dhawal, who will talk us through AI security and data security. And that's an area I know many of you would like to know more about, so listen up. Then Dhawal will host a panel with 3 customers that we're very grateful to have with us today. They'll be sharing their security challenges and their journey with Zscaler. They'll also be available to take your questions. So start thinking about what you might want to ask them.

And then after a short break, our new CRO, Ross Tackett, will introduce himself to all of you and share our go-to-market strategy, and then Kevin will finish us up with some financial perspectives. And then we'll welcome all the speakers back on stage and take all of your questions. And for those of you here with us in New York City, after that, we'll invite you to join us for lunch and a reception, and we also have some demos of the products that we'll be sharing today. So make sure you hang around for that.

So now the part that everyone loves, the forward-looking statements. Before we begin, I'd like to remind everyone that today's presentation contains forward-looking statements within the meaning of the safe harbor provisions of the federal securities laws, including statements regarding our future financial performance, business strategy and market opportunities. These statements are subject to risks and uncertainties that could cause actual results to differ materially from those projected, and we undertake no obligation to update them. For a more complete discussion of the factors that could affect our results, please refer to the risk factors described in our most recent filings with the SEC, including our annual report on Form 10-K and quarterly reports on Form 10-Q.

And with that, it's my pleasure to introduce Zscaler's CEO, Chairman and Founder, Jay Chaudhry. Jay?

Good morning, everyone. Great. Thank you all for joining us at this Zscaler's 2026 Investor Day.

As you heard from Kim, we have a great lineup planned for you this morning, and I'm excited to hear from leaders across our business. Overall, we expect you to leave today with 3 key takeaways. Number one, Zscaler is the cybersecurity platform for the AI era, and AI is the most powerful tailwind of driving our business. Number two, we are capitalizing on this opportunity by expanding our platform and strengthening our go-to-market muscle. And finally, we have multiple growth engines that we expect to drive our reacceleration in the years ahead.

So let's get started. Zscaler takes me back to 2008, when I started the company, when we pioneered Zero Trust security. The term Zero Trust wasn't well known. But what we came up with the idea was trust no one. And making the Internet and advanced technology safer to do business was our primary focus. We weren't focused on building the next and next-gen firewall. And with this, our vision was ushering in a world where people, devices and AI agents can communicate and innovate securely. So security shouldn't be holding them back. And we achieved a lot in the past some 18 years or so. Look at the numbers, over 50% of Fortune 500 companies, 750 billion daily in-line transactions we inspect and enforce policy. Over 200,000 locations, these are public globally distributed, and there are many more private ones. And over 1,000 patents issued and granted. Lots of cool technology. We are essentially about innovation.

So the new approach to security we took was driven by the fact that in my background, where I spent over a dozen years before starting Zscaler doing security. And I've seen more and more security products coming and more and more breaches happening. I was convinced that the traditional way of doing security by building these moats with firewalls wasn't going to work. So the new approach was Entity A should be able to talk to Entity B through our exchange, our switchboard, literally like a phone switchboard in the old days. In the old days, when you called, you said, I want to talk to John, you got connected to John and John only. Then you said, I want to talk to the President of the U.S. They said, "Sorry, go away, you're not allowed to."

Simple concept. Party A can only talk to Party B as opposed to the old concept. Everyone is in. Now let me try to put the guards and one firewall here and firewall there. That was the simple concept, the simple idea. And the goal was fundamentally twofold. One was stopping threats, no matter where they come from. Two was protecting data. Two basic things. While we started with users, the architecture designed to handle any entity. Today, we handle devices, sitting in plants and factories. We handle all kind of AI agents, the most exciting part. We've taken it to cloud workloads and third parties and the like. And they should be able to access the right data sources. There could be AI models, could be applications, SaaS applications, agents and whatever tools. It's a very exciting platform, and it is probably far more needed today than it was ever before.

Now I'm going to talk about why Zscaler is winning the business and why we have a durable growth opportunity at Zscaler. Four simple areas. Architecture matters, and you can't fix it overnight. A unified integrated platform takes time to build. You can't buy a bunch of companies and call it a platform. Removing cost and complexity. And lastly, our go-to-market engine.

Let me walk you through each of the 4 areas. The first is our architecture, okay? Many times, people get confused between architecture and features. Architecture is the foundation. Features are you can add on. You don't build foundation every day, okay? Now our story starts with building disruptive solution, disruptive technology. I've always liked not to build me too, but a different better way of doing it. Mobility played an important role in Zscaler's success. That is a big mega wave. Cloud was the second mega wave that helped us because in the world of mobility and cloud, the old security based on firewalls and VPNs made no sense. We leveraged those 2 [ gig ] mega waves. And now we see the AI era, I call it a giga wave because it's going to change things. It's already changing things. And AI is the largest tailwind we have ever witnessed. We are already seeing early signs of it. And we think there's a lot more to be coming.

Now AI. There's hardly a meeting that goes by where AI is not part of the discussion. I do lots and lots of customer meetings. Last week, I was in Dallas, meeting 80 CISOs of large enterprises. Every discussion starts with, my CEO, my Board is asking me to really roll out AI solutions at a faster and faster pace because CEOs are convinced, and rightfully so, that it can accelerate the top line growth, and it can also drive efficiencies. But they're all worried about some of the challenges, especially the cyber issues they have to deal with.

I'm going to talk about how we deal with, how we help our customers with cyber risks to manage them so they can effectively start rolling out solutions. And there's cost and complexity of the current infrastructure they need to deal with on top of which rolling out AI becomes very, very hard.

So the tailwinds for Zscaler platform can be put in 2 buckets. One is frontier AI attacks. These models are powerful. These models are dangerous. And we'll talk about how we empower our customers while mitigating the risk. And the second part is AI agents. Agents we deploy or agents others deploy, agents going rogue, how do you use agents effectively, okay?

Let's start with the risks posed by frontier AI. For that, one needs to understand how most of the world has worked in the past. The way technology is done is, I'm going to build my application portal, my mobile portal and put a VPN, firewall, load balancers in front of it, and it's all open to the Internet so people can come over the Internet and connect with them. That was good. But the more things exposed to Internet, the bigger the attack surface. We had some amount of attack surface before AI. Now AI is further increasing the attack surface. You're adding AI chatbots. You got MCP servers, you got AI agents. The attack surface is getting bigger. We call it attack surface because anyone from the Internet can scan you and find you.

That was there before. But now once they can find you with AI models, they can find all kind of security vulnerabilities. It's unbelievable how many vulnerabilities can be found. Once they find them, they compromise it and then the game gets over. That's how a lot of these breaches are happening. So the risk of frontier model is exploiting these vulnerabilities and the gap between discovery to exploitation, which used to be months is now becoming minutes or seconds. That needs to be handled and you can't fix it by putting on more DDoS protection, more firewalls, more VPNs.

Here is a real-life example of a customer. It's an energy company. Rather than reading all the stats, I'm going to highlight 3 things: Internet-facing assets. These are exposed infrastructure, your applications, your firewalls, your VPNs, your load balancers and all those things, about 6,614. It's not an unusual number. That's what we see in large enterprises. And for that enterprise, 1,049 software vulnerabilities. And one thing worth highlighting, VPN appliances, 40 of them. It's easy to find every other week, you see some VPN got compromised. This is the landscape you're dealing with and models, these frontier models can exploit them very easily.

Now how do you fix them? Not buying more of those things, but simply using a different architecture. In the Zscaler world, all applications that require authentication are hidden behind Zscaler and attackers can't breach if they can't reach. Very simple, elegant architecture. That's what our customers get very proud of because simpler but great solutions.

The second area is the lateral movement of threats. Once they compromise a firewall or VPN or application portal, the malware moves laterally. How does that work? I tried to explain that concept. Generally, it gets confusing between networks and all. So I came with a new idea. Let's use the highway metaphor. Just like highway connects different cities, the corporate network, the private network connects your data centers, your SaaS application, public cloud, your new clouds and all the branches and everything. The goal is once you get on the corporate network, you move left, you move right, that's a part of the job of the network. That's wonderful.

But now we got, first of all, more applications. On top of what we had, we got AI applications, we got agents and all that stuff happening. And what can bad guys do? They can get on your corporate network, they can move laterally, they can reach targets and cause breaches. It used to be hackers before. Now you're going to see more and more AI agents acting autonomously being able to do this job. That's a big risk that needs to be dealt with.

Let me give you an example of real life. We have been reading about Hugging Face and OpenAI agent. How did it happen? Well, OpenAI had this agent in a sandbox. It got out of the sandbox. It got on the network. It moved laterally, then it got to found a way to the Internet. And then scanned, it found Hugging Face website, then it discovered vulnerabilities in the website and then it breached it. Simple concept, but this is a great example of how breaches happen. I wonder why people are surprised about it. This has to change. Architecture has to change. And that architecture is going to help us.

Now one other point. Today, we worry about users being the weakest link. Tomorrow, agents will be the weakest link and far more dangerous than people because they can move at machine speed. They know they need no coffee break, no weekend, no evenings, no vacation, and the numbers keep on growing. That's the big risk. How do you fix it? By taking -- adopting a different approach. That's where Zero Trust comes in. Remember, I said in Zero Trust architecture, entity A talks to entity B through a switchboard to an exchange. That's a simple policy engine. The difference, the old approach was you put everyone on the network, now you try to kind of control who goes where. This is the difference. That's why you can't build it on top of traditional security. You can add -- you can't bolt on Zero Trust on firewalls and VPNs.

Let's move to the second area, and this is our comprehensive and unified platform. We started Zscaler with securing users. First of all, Internet, second for private applications that you knew as ZIA and ZPA. In '18, when we went public, we had just one product, ZIA. And investor used to ask, would you really be able to go beyond ZIA? Look at the journey we have come along. We built full Zero Trust for users with Internet access, private access and user experience. Then we added a branch because branch and devices must be Zero Trust. The big risk out there is these networks that are like a mesh network, everything connects to everything. Whether you have MPLS or SD-WAN, you have risk of lateral movement.

We brought the solution where each branch is simply like an Internet cafe. It's an island of its own. Each device in a factory or a plant is isolated. It can't infect each other, very powerful technology. I told you years ago that MPLS had to go away because with Zscaler architecture, everything will go direct over the Internet. People kind of said that was a crazy idea. I can tell you, it's a matter of time, you'll find that SD-WAN used to be something. Today, when people talk about Zscaler's competition, they compare us to SD-WANs of the world out there. I really will show you how our customers are embracing Zero Trust branch, and you hear from one of our customers about this story where they've done a large-scale transformation from the old world.

Zero Trust Cloud, more and more workloads and applications are being built in the cloud. And how is security done in the cloud? Still the old way. Just like people used to do application lift and shift, and then they realize that lift and shift wasn't a good thing. Now they build cloud-native applications.

But they're still lifting and shifting security. You know how? Firewall boxes for East, West in the data center. Now it's virtual firewall in the cloud. Firewall is an IP device. It's a network device that deals with IP addresses. This source IP can only talk to this destination IP. That is hard enough to deal in the data center. Think about dealing with lots of moving IPs when the workloads are ephemeral. How do you deal with that? Impossible. The best solution would be Zero Trust. You can say, workloads in VPC A can only talk to workload VPC B. Powerful concept. Workloads with Tag A can only talk to workloads with Tag B. Those are the innovations we bring to the table. That's why our customers are so excited about working with us.

And then the most exciting area of expansion of Zero Trust journey is AI agents. Dhawal is going to talk about how we really built. We have the underlying core, most of the technologies in place. We had to build a number of things. We had to build MCP gateways, A2A gateways. How do you extract and analyze prompt? How do you figure intent? Those are the type of things we build. But we're already sitting in the traffic path. We already distributed infrastructure. So we believe that we are in the best position to really provide the solution for controlled managed use of AI [ agents ].

So this brings together to the overall platform that's integrated and comprehensive. Zero Trust SASE everywhere. Users to branches to cloud to AI agents. Then on top of that, data security is fundamental, extremely important. Our customers want data security that's integrated 1 policy rather than 5 different data security products. In this area, we have very strong advantage. We have over $410 million ARR. Probably, if this was an independent company, it will probably be the largest data security company. And we are the natural party to do data security because almost all data leaks to the Internet. We're sitting in the traffic path for that.

Security for AI is the next big opportunity. If you're deploying AI, how do you secure it? And Dhawal will get you deeper into that. And the other new area we recently added is agentic SecOps. SecOps is ready for disruption. Adam will take you deeper into it. But one comment I'll make is we have better, more meaningful data for SecOps than anyone else out there. And two, anyone who started building SecOps before a couple of years ago, built it the old way. They may be using AI, but they were using AI to supplement something built for humans. We actually built SecOps for AI agents for the native use.

Let's -- so what is our unique advantage for architecture and platform? Our scale. Scale matters, scale will become bigger and bigger with AI. Cyber better protection because of an architecture that gives us minimal attack surface and prevents lateral movement. A platform that covers everything for Zero Trust everywhere is number three. Number four is our securing AI and data everywhere. It's hard to find someone who really can provide this security everywhere. And lastly, unmatched proprietary data. AI models are getting commoditized. There will be special models and open [ web ], all this stuff, but private data will matter, and that's the unique and lasting advantage we have.

Let's move on to the next topic, cost and complexity. If you look at a typical enterprise, this is what their infrastructure looks like. Everything connects to everything because they want to reach everything and trying to secure a firewall here, router here, switch here, all that stuff is very complex. It's hard to find a CIO and Head of Infra that says, my infrastructure is simple. And then they keep on adding more and more products on top of that. It's really a mess. And what does it result in? Well, you got high cost, CapEx and OpEx both. Scaling challenges is a problem. Complexity. Complexity is the enemy of resilience and complexity is the enemy of security. And those things need to be fixed. That's why you couldn't fix it by bolting on more and more features on an architecture that's 30 years old. You have to do it in a new way. And this is where our Zero Trust architecture comes in.

One other clarification. The word platform has meaning -- lost meaning out there. Every company claims to do a platform. Platform actually uses a common set of services for policies, some of the other stuff. That's a real platform. I see many customers talk about and say they got sold a platform. When they looked underneath, they found that it is 8 policy engines, 8 products. That's a real quote from large retail in the U.S. who I met with about 2 years ago. We are very mindful of building a right platform. That's why even when we do an acquisition, it's generally a smaller tuck-in company, so we could integrate the solution and make it part of our platform. And if we do it right, we're going to -- we bring a lot of cost down.

In this simple example, the various areas, even if you exclude legacy sunk cost because that's already paid for and the customer wanted to say, let's deal with that in Phase 2. We're able to bring the cost down significantly. Typically, 30%, 40% range is not unusual to take out. And that's because we don't worry about the cannibalization of firewalls and VPNs and load balancers and those express routes and [ NVA ], like all those things go away with Zscaler. Firewall companies have to protect the installed base because if that goes out, their business hurts. That's why we can take a lot of this stuff out that others won't do.

Let's move on to the fourth topic of why we win. It's our go-to-market engine, okay? our TAM, you've seen the numbers. It started with 0 to the users at about $62 billion. It's moved up at $220 billion because of all these platform offerings were added. We are dealing -- and these are very logical and systematic plans. We don't just go and randomly buy a product here, buy a product here. These are very synergistic areas, very well thought through. We're not trying to be everything for everyone, okay? And I'll highlight that in the previous chart, I showed our AI security TAM of $24 billion. That's about 1% of the total TAM of AI spend that Gartner is predicting. Most of the conversations I have with CIOs, they tell me that they're factoring mid-single digits of the total AI spend for AI security. If I look at that 4% or 7% range, that will be in the $100 billion to $200 billion range. So we kind of left it outside. But just to give you a data point where the numbers come from.

How is AI helping you? Let me give you -- be more specific. First of all, these tailwinds are coming from first 2 areas, large attack surface and increased lateral threats. Those 2 things are bound to bring more breaches. We're already beginning early signs of that, which really means customers are appreciating more and more need for Zero Trust architecture. And this will give us opportunity to sell new logos as well as upsell. And second area, we're beginning to see there's more traffic. There are more entities. There are billions of agents. Communication, when we secure communication, we're dealing with more traffic. That means exponential traffic growth, which will drive security consumption, which will drive upsell opportunities for us.

Let me give you some real data points. Since Mythos' preview program, we've done over 400 assessments with a large Global 2000 customers. And what are we finding? Well, as I talked earlier, large installed base that needs to be minimized. Internet exposure need to be minimized, and that drives need for Zero Trust private access. Lateral [ threat ] movement, which typical networking brings. We need to eliminate trusted network. Network should merely with the transport and plumbing. It should never be trusted. Zero Trust SASE everywhere is a solution for that. Attackers are scanning the targets once they get on the Internet, once they get on your data center. Maybe in the data center, they get on the network, maybe in the data center, maybe in the cloud.

We have honeypot technology deception to handle that. We are seeing meaningful increase in that demand for it. And -- if breaches happen, then they take your data loss. So there's a bigger demand for data security. And lastly, AI is showing up, shadow IT agents showing up. There's a lot of demand we are seeing at a pretty rapid pace with short sales cycles in this area. So these are tangible examples of the AI risks that are leading into growth opportunities.

The 4 levers of GTM that Ross is going to expand upon. Platform, you heard the expansion story. New logo engines with the platform expanded. We have many, many areas to land on. There's no -- we don't need to go through just one traditional way of Zero Trust for users. You can start from AI side, you can start for data security side and others. Partner leverage, we spent a lot of time building our partnerships. Global system integrators are especially useful because they actually help drive the transformation and customer obsession.

Our customers are very proud of what we do, or what they do using Zscaler. You'll hear from a few of them today. But here is the stat I'm very proud of. This is based on some data we analyzed. This is about 9- or 12-month-old data, 285 CXOs, this large company. They bought Zscaler in 2 companies, went from company A to company B, called us, 84 of them in 3 companies and 45 of them in 4 companies. Very, very wonderful to see that.

So our financials, strong financials. You know the numbers, $3.8 billion ARR. That's about 25% year-over-year growth. Revenue growth in the same range. Net new ARR, which is an indicator of near term how things are happening, you're seeing some very good results. Look, we started at 7% in fiscal '25, took it to 14%. Our first half of fiscal '26 was 10%. Q4 was 17%. It's moving in the right direction. We are going to keep it moving in the right direction, and there's enough opportunities to do that. And we also managed our bottom line very well. Kevin will give you more about that.

Here's what I'm excited about, the big opportunity for us with all that stuff. We think we can take -- get to $8 billion to $10 billion by fiscal '31, okay? And there's a lot of opportunities, the reasons I already stated to be able to drive 17% CAGR on the base level and 22% CAGR is very much achievable. That's a potential upside for us. Very excited to build the business to take to those levels, okay?

Why invest in Zscaler? If I boil it down to 4 simple reasons, I stated enough about the architectural advantages. Others can't just copy architecture. It's the foundation you have to do it right and you have to worry about cannibalization. That's why you aren't seeing firewall companies building on Zero Trust. They may claim Zero Trust once in a while, but it's not there. The demand tailwind I talked about untapped go-to-market. There's a lot of opportunities in the customer base as well as new logos for us to do and very strong financials.

So what are we going to hear today? So for the rest of the presentations, right, our team will give you a deeper view of it, okay? Platform expansion, Adam and Dhawal will share more detail about the platform and the recent innovation we have done. You'll have a chance to directly hear from some of our customers about this area, okay? And then Ross will take you through our go-to-market engine and the steps we are taking to drive growth moving forward. And finally, Kevin will share some financials, and should give you more confidence in our path to $10 billion ARR and provide a fresh look at our long-term financial model.

So let's get started. I'm thrilled to welcome Adam Geller to stage. Adam?

All right. Thank you, Jay. Welcome, everyone, here in New York, everyone on the webcast with us. I'm excited to spend my time talking to you about all the new innovations we have in our platform, and I want to connect those to how we're enabling new growth opportunities for the Zscaler business.

So -- we've been relentlessly innovating on our platform for the past 17 years to solve some of our customers' biggest cybersecurity challenges. This is leading to our key Zero Trust solutions, and we're delivering them for some of the world's largest organizations. We started with Zero Trust for users well before SASE existed as a market category. And from there, we extended the principles of Zero Trust to branches, to OT, to IoT, and now we deliver Zero Trust SASE everywhere. And it's this foundational platform that's also enabled us to deliver new adjacent solutions over time.

So for example, we launched data security, and we started with our in-line data loss protection, or DLP. And we've since expanded that over the years to cover all of the important channels for DLP. So that's endpoint, that's data at rest, data in motion, data in the cloud, data in SaaS applications, even data on-premises as well. And next, we delivered security for AI. And this is a space where we've innovated very rapidly over the past several years. And Dhawal is going to dive deeper into this right after my presentation. And most recently for us is agentic SecOps. And we just launched the latest update to our agentic SecOps offering last month. And we're bringing together -- to do that, we brought together our strong agentic SOC product capabilities with our expert-led services around managed detection and response and managed threat hunting.

Now Jay mentioned this, the rise of AI-driven threats and the massive increase in vulnerabilities discovered by frontier AI models like Mythos has created a huge uptick in demand for Zero Trust solutions. And what that means is Zero Trust is no longer just a strategic security transformation for an enterprise. It's kind of changed. It's now become an urgent board-level mandate to help organizations protect themselves against AI-powered attacks, but also to help them safely enable the adoption of AI for the organization itself.

The fundamental truth that existed, and this existed, by the way, pre-AI, it's now more true than ever. If you are reachable, you're breachable. And this continues to be true for users and applications, and now it's also true for AI and for agents. The reality is Zscaler, as a company and as a product, it was built for this moment. So it turns out the unique architecture, Zero Trust architecture from Zscaler that our customers have used so effectively to secure their users, it's also the best way to secure AI agents and to protect against AI-driven threats. We continue to expand our platform on the foundation of Zero Trust Exchange to solve the broadest set of cybersecurity challenges for our customers.

Now we're going to share more details about our expanded set of capabilities beyond Zero Trust for users and how this has enabled us to address even larger total addressable markets. But first, I want to come back to this point that Jay mentioned again, which is why does the architecture really, really matter? So you may think of Zscaler as a network security company. We have a fundamentally different point of view on that. So networks, whether it's a LAN, WAN, SD-WAN, VPN, whatever technology you're talking about, their whole goal is to connect things together. Zero Trust is not about connecting things together, right? So earlier, Jay touched on our approach to securing the agentic enterprise. And it's based around this premise and this principle here that secures any to any communication using business policies and not networks.

So let's dive a little bit deeper into that. Our fundamental approach to security lies on abstracting away the network completely, right? That's the primary job of our Zero Trust Exchange. So to us, that network, it's simply plumbing, and we never trust them. So that means you never put the user, the device or the agents directly onto the network. Instead, that's where we use the Zero Trust framework to connect entities only to the resources that they should be allowed to have access to.

So let's take that one step further. How do we do this? When an entity tries to connect to resource, our Zero Trust Exchange is going to quickly run through a process in real time to do the following things. First, we're going to verify the user next or the entity. Second, we're going to establish what should they be allowed to do. Then we're going to bring together first- and third-party signals, context signals because we want to understand the user, the entity, the device as well as the potential risk in whatever interaction is about to happen. And then finally, because we're in line, we're able to enforce a policy decision on what actions to allow and what actions to block or to deny.

Now the net result of this is a dramatically more secure environment that can protect against the 4 stages of an attack. So the first step is, number one, we'll minimize that attack surface. This is by hiding applications behind our Zero Trust Exchange. This way attackers can't even discover them. Second, if an attacker does somehow manage to get in, we prevent breaches, systems from being breached with our in-line cyber protection capabilities. And then third, we're going to prevent attackers from moving laterally across the environment where they're looking for those high-value assets. With the Zero Trust model in place, everyone's untrusted, every device is isolated. That's how Zscaler customers operate their branches, their factories and their cloud environments. And then fourth, we're going to use our in-line protection capabilities, data protection capabilities across all channels to make sure that we can stop sensitive data from either being accidentally or intentionally exfiltrated out of the organization.

So why does this architecture matter? Not only do we connect users and agents only to what they need, unlike competitive solutions where the users and agents land on the network, but we also sit inside every traffic flow where most other solutions observe that traffic from the outside. We are a single unified platform where others stitch together different point products. And in an AI world, where exposed applications can easily be exploited, this difference absolutely matters. As I said earlier, if it's reachable, it's breachable. And we stop attackers from even being able to reach the applications in the first place, which does wonders for preventing the possibility of a breach.

So while we continue to drive innovation across -- and growth across our platform, a big focus in our product development over the last 12 to 18 months has been building new product beachheads. This is an opportunity to land new customers and to expand to our existing customers as well. And we see this in 3 main areas: users, branch and cloud. So let me walk you through what we're doing in each of these areas, and I'm going to start with Zero Trust for users.

So Zscaler Internet Access, Zscaler Private Access, so ZIA and ZPA, they have been the foundation of our platform for the longest time. This is how we secure users accessing the Internet, SaaS applications, private applications, whether they're hosted in customers' data centers or they're hosted in the cloud. We've been very successful helping customers understand how to implement Zero Trust for users to transform their overall security architecture. So this is a strong part of our business, and it's growing in the double digits. And these transformation projects, they are -- when they complete, they take Zscaler and they put us in place as mission-critical service for our customers. And the scale of this is unmatched. So Zscaler now protects nearly 7 million business applications for our customers and safely enables more than 55 million users every single day. Zero Trust transformation for users with ZIA and ZPA, that's typically how customers have started their journey with Zscaler, and we've continued to drive industry innovation, specifically in this space.

Now in particular, let's look at ZPA for a minute. It's seen a sharp increase in attention from customers in 2026 from the tailwinds created by AI. Mythos-like frontier AI models and AI-powered attacks are finding vulnerabilities, they're chaining them together to evade legacy perimeter security solutions. With ZPA, we're known for hiding that attack surface area and preventing lateral movement. We also now protect applications at run time with our new Autonomous AppShield, virtual patching. And so this solution uses frontier AI models to continuously scan these applications that we're protecting. It's going to find vulnerabilities, create virtual patches for them, and we'll be able to protect these applications before a customer even has a chance to patch them. Just yesterday, we announced our partnership with IBM and Red Hat to deliver virtual patches for vulnerabilities even more quickly, often before they're publicly disclosed.

So ZPA represents a meaningful expansion to our installed base alone as approximately 30% of ZIA user licenses today don't also have ZPA. So it also drives new logo opportunities as prospective customers are urgently looking for new solutions to protect themselves against AI threats. Earlier, Jay showed a slide and he talked about the Frontier AI model assessments that we did with over 400 organizations. The #1 ask that comes out of those assessments is, what do I do next? This is exactly what we tell them to do. And a Global 2000 financial services customer was looking to reduce their exposure to vulnerabilities found by Mythos-like AI models. And they expanded their ZPA footprint to cover all 120,000 of their users. This increased that customer's ARR to Zscaler by nearly 50%.

Okay. Now we're going to move over, and we're going to talk about browser security. The browser is where modern work gets done, but web content can be risky. It can also provide -- also providing third-party contractors access with BYOD, so bring your own device. It's historically been hard and complicated. You think about it, a dangerous website, it can serve malware down and compromise an end user's machine. Also limited controls can allow sensitive intellectual property that's in that application you're trying to protect to be extracted out. In Zscaler, we pioneered one of the most effective solutions for this problem, it's called cloud browser isolation. And so with this approach, user accessing an application or a potentially risky website, they only see a streaming image of what they're browsing. And this means that they are protected against malicious content because no content is coming down, and the organization that's controlling this can easily decide what sensitive data even gets delivered to the user, what can they see and what can they do with it.

Now while full cloud browser isolation is arguably the most secure approach out there, there are some cases like with third-party contractors where a simplified user experience through a local browser might be preferred. And so over the past 18 months, we've expanded our browser capabilities to include a browser extension that lets a customer work securely in any browser that they want as well as building and releasing a dedicated enterprise browser. And so with this, Zscaler now has the most comprehensive browser security offering in the market. And we think it can satisfy customers across any use case and beat any of the point product competitors.

The result of this for Zscaler is a net new land opportunity where browser security starts as a potential first step in a customer Zero Trust journey, and we saw this play out with a Fortune 500 retailer, where we won a competitive deal against other browser-only security vendors. The customer who bought Zero Trust browser for 350,000 of their users, they had no previous Zscaler deployment, no ZIA, no ZPA. This was the land for it. And then they plan to expand to more parts of our portfolio in the coming quarters.

So that's just some of the incredible innovation we're driving in Zero Trust for users. Now let's talk about branch. Legacy branch solutions like SD-WAN, VPN, MPLS, they're typically quite expensive. They can be complex to manage, and they drastically expand the overall attack surface for a customer. Remember, all these are networking technologies. So they were built for a purpose, and that was to connect things together, not to provide security. We introduced our Branch Connector several years ago, and that was our initial solution to this problem. Over the past 12 months, we thoughtfully unified our Branch Connector and our microsegmentation capabilities into a line of purpose-built appliances that provide a true secure Zero Trust alternative to SD-WAN.

The unified appliance that we built, it seamlessly connects the branch location to our Zero Trust Exchange and then leverages local microsegmentation to eliminate lateral movement within that branch location or factory. And so just by way of definition, microsegmentation, it's that ability to place every single device, even small footprint IoT, OT devices into isolated segments. So this is where they can't communicate with any other device unless we explicitly or the customer's policy explicitly allows it. Remember that these devices, for a lot of customers, especially in factories, they're in places where they can't install additional software or software agents onto it. So it's critical that they're able to easily segment out their infrastructure to prevent lateral movement in these environments.

The result again with branch security is a new land-and-expand opportunity for Zscaler. Highlighting this, it's a recent deal we won with the Global 2000 manufacturer who's deploying Zero Trust branch to secure their OT environment. Again, this customer had no previous Zscaler deployment. Their incumbent network provider had a shot at winning the deal to secure this environment, but the customer chose Zscaler to secure all of their critical production sites because of the way we approach this in a Zero Trust fashion.

Now 40% of Zero Trust branch customers are new logos to Zscaler. So this is a great opportunity for us to accelerate our new logo growth. And again, it also provides an expansion opportunity to the rest of our customers once they are on the platform and are going down the Zero Trust journey with us.

So now on to cloud. In the cloud area, enterprises have been struggling to secure cloud applications running in hyperscalers like AWS, Google, Azure, OCI. And the traditional approach is to bring that what you know, so bring legacy firewall appliances and move them into the cloud, and they can be costly, complex to manage and not often able to really stop the lateral movement that we would want them to protect against. So when we started in cloud security space, we started with a Cloud Connector. This made it easy to secure cloud workloads running in public cloud. And customers liked that simplicity.

They liked that flexibility. And they really decided they wanted this to be extended more and to be able to replace all the different use cases that they would want, where they would typically need to deploy a virtual firewall. This is why we introduced our Zero Trust Cloud Gateway in late 2025. And this is a managed solution designed to deliver Zero Trust for cloud workloads. And it includes microsegmentation. And overall, we'll run this service on behalf of our customers, so there's no need to deploy and manage virtual firewall infrastructure.

If you remember just before, I talked about this idea of microsegmentation in the branch, and the goal of that is to stop lateral movement. We apply that same principle natively for cloud workloads. This time, it's an agent-based approach. And that's what allows us to do full -- all directions, north, south, east, west traffic flows in cloud environments in a true Zero Trust fashion. This, too, creates new land and expand opportunities for Zscaler. We saw this at a global manufacturer. They deployed Zero Trust Cloud in a 7-figure deal to implement microsegmentation. The goal was to eliminate virtual firewalls, and they did this following a major cybersecurity incident, which brought their physical production environments down for over 6 weeks. This customer was able to get our Zero Trust cloud solution up and running in less than 10 minutes during their proof-of-value process, and that demonstrated the simplicity of our powerful cloud solution.

So Zero Trust, as we described, this is a journey, right? And as I've described today, customers are starting that journey wherever it makes most sense -- the most sense for them, could be users, could be branch, could be cloud and increasingly, it's going to be AI agents, which Dhawal is going to talk about in just a moment. Any of these are starting points for customers to adopt Zero Trust SASE everywhere.

And with our relentless innovation, our proven success in the market, we are really proud that Zscaler has been recognized as a leader in the 2026 Gartner MQ, Magic Quadrant, for SASE platforms. And if you dig a little deeper into that, Gartner says, 50% of new SASE deployments are going to be based on single vendor SASE platforms by 2028, and that's up from 30% last year in 2025.

So with what I described to you with Zscaler's Zero Trust SASE everywhere, we think we are super well positioned in this space to win with a comprehensive single-vendor platform in the SASE space. So we continue to expand our Zero Trust SASE Everywhere cloud presence too, to support more and more customers across more and more locations. And now what's increasingly important in many parts of the world now is cloud sovereignty, right? Our leadership in sovereign cloud, this is a structural advantage for Zscaler. The nature of our architecture, it provides us with the ability to maintain separate traffic processing, management and log storage, and we can do this in isolated regions around the world. And as digital nationalism rises globally, Zscaler is the only cloud-native platform that delivers global performance while we're meeting strict local data residency and unique regional requirements.

So for example, our partnerships with providers like Schwarz Digits, who combine our Zscaler Zero Trust Exchange with their European sovereign cloud STACKIT, they continue to support customers with sovereign cloud requirements, or together, we do, in ways that our competitors cannot easily serve. You're going to see us make further sovereign cloud announcements in additional regions over the coming year.

So in my last few minutes, I want to talk to you a little bit about agentic security operations or Agentic SecOps. Now along with availability, reliability, performance, our customers benefit from our unrivaled telemetry across identity, network, endpoint, cloud, and they get this when they leverage our in-line Zero Trust Exchange. Now it's important to note, without accurate and relevant data, the value of applying AI models, especially in the cyber world, it's greatly diminished. But with this unique foothold we have on the endpoint as well as with this global cloud presence, we can see and stop threats that others are going to miss. We believe one of the most important ways to leverage these powerful vantage points and this telemetry is to apply it towards detecting and responding to cyber-attacks. This is why we launched our agentic security operations solution a few weeks ago.

So let me introduce you to our Agentic SecOps offering. Legacy SOCs, security operation centers, they were typically built to chase endpoint alerts and to search SIEM logs and do this after the fact and most typically with humans in that process to try to figure out what happened after a threat was detected. Now that model itself has struggled in the past and now it's completely broken with AI-driven threats that are operating at machine speed.

So with Zscaler's agentic SecOps, we started with an agent-first approach. We have over 50 agents that are trained on our global telemetry and they're tuned by 10-plus years of human security experience that comes from our Zscaler ThreatLabz team as well as our Red Canary Managed Detection & Response team. In our agentic SecOps platform, we also brought together exposure management and threat management into one place. So what does this mean? This means that a customer can proactively work to reduce their vulnerabilities, their exposures and their risks and then they can use that information to quickly contextualize and automatically respond when a security incident is happening. That information is super valuable to go through an investigation. And so time and time again, our telemetry has proven to be very valuable to our customers and helping them investigate security incidents. And now we're operationalizing that value for customers with our agentic SecOps solution.

SecOps, too, does provide yet another land-and-expand opportunity for Zscaler. We had a Global 2000 manufacturing company engaged us to help quantify and manage exposure and risk across 700,000 assets. That was before Mythos. When Mythos came through, this customer realized that sampling their estate is just it was not going to be enough. They adopted our agentic SecOps solution and expanded their coverage to more than 20 million assets to make sure that they could cover their entire organizational footprint. Now we're in the early innings of reimagining security operations, but we are really excited about our unique opportunity in this space.

So to summarize, the Zscaler Zero Trust Exchange, it is the platform for the AI era. We think it's the best way to secure AI agents and to protect against AI-powered attacks. We've continued to drive innovation across our platform to expand our TAM. So not only does Zero Trust for users remain a powerful growth engine for us, but we've also unlocked new and upsell opportunities across users, branches and cloud. In addition, we're excited to disrupt the SecOps market and transform how organizations detect and respond to threats, which we believe will open up a large adjacent growth opportunity for Zscaler.

And with that, I'd like to thank you for your time today. And now I'd like to welcome Dhawal Sharma to the stage, and he's going to dive deeper into what we're doing in AI and data security. Thank you.

Welcome, and thank you, everyone, for joining us here. In this section, we're going to talk about two broad areas of our platform, security of AI and data security, which unlocks a significant TAM for us, as Jay showed, which is about $48 billion. And we have a sizable book of business already where we are securing data for our customers over the last many years. And earlier this year, we also introduced our AI security portfolio.

As we have been selling this in the market, what we are seeing is there is a strong synergy between the two. The persona of Chief Data Officers is overlapping and merging with the Chief AI Strategy Officers, and CISOs are getting deeply engaged in securing AI everywhere. Also, when you think about the traffic patterns, 80% of unauthorized AI traffic caused a data security violation. And as AI creates more data, organizations are leaning to build safeguards against AI with Zero Trust. And these are strong statements that have come from Gartner recently, and we do strongly validate that in our customer conversation as well.

As we think about AI security, we are seeing AI actually creating strong tailwinds for data security as well. We are seeing that AI adoption is growing in organizations. We have a strong book of business of more than $400 million in data security today, growing with a CAGR of 31% as well as, in the last 3 years, we have seen strong growth of 126% in this area. What is even more exciting, however, is that AI security creates a new entry point of Zscaler for customer acquisition. What I'm very excited about is the fact that 70% of security for AI deals in our fiscal Q4 actually had data security with it. And as we think about these 2 problem statements, they are actually converging and are becoming very symbiotic in nature.

To understand how these 2 areas are coming together, let's look at the underlying problem statement. So in today's world, every app we have built over the years, every SaaS application is becoming agentic and is creating more data. Fundamentally, it is becoming harder to understand who has access to data, what they are doing with it and what is the risk of their access to the data? Moreover, AI models are not just consuming your enterprise PII data and transforming that data. They are also creating more data. This creates more data attack surface for the organizations and also AI agents can work autonomously with no human in the loop, exfiltrating enterprise data to risky destinations. In my observation, this is a new frontier that enterprises have not tackled before.

Zscaler has been working in the area of AI security for many years. So let's talk about our security for AI solution for a second. We have been building this for more than 4 years, and you think about the moment when ChatGPT came out, right? Since then, there has been explosion of generative AI and agentic AI application. There are thousands of them that have popped up since then. We introduced our gen AI security product that provides visibility into shadow AI and also secures thousands of gen AI applications. Then we introduced AI Guard, one of the industry's first intent-based policy framework to secure what happens inside the prompts of these applications.

Last 1 year, we are seeing that AI adoption, especially within the hyperscalers and new clouds has accelerated significantly, which means people are building more AI infrastructure in the cloud. We introduced our AI asset management solution then to bring visibility into AI everywhere and then secure the posture and configuration of these AI applications as well.

Then we also introduced the ability to do offensive red teaming through our acquisition of SPLX on AI applications that organizations are building to determine the risk of AI apps. 2026 is the year when agents are going mainstream, whether embedded inside SaaS applications, whether they are built independently for the organizations to understand the relationship of AI agents, data that they've been using and the identities that they're using, whether they are human identities or nonhuman identities, we introduced the AI access draft through our acquisition of Symmetry Systems. We also introduced endpoint AI security that is built using our acquisition of SquareX. And finally, we introduced our AI gateway that sits between every AI interaction within an enterprise and secures them with -- and we'll talk more about it soon.

When we think about security for AI and data security, there are 3 big pillars in this area as we bring these together. And the thing that really stands out for our customer is that when we talk about AI, it also includes how data is coming with it. So the 3 pillars are discovering data everywhere and making sure the AI that is using that data is also discovered, securing every AI interaction and securing it with against cyber incidents like prompt injections and the data that is getting transferred in those interactions and protecting those AI apps and infrastructure that you are building. So let's go deeper into the discovery to understand the shadow AI usage, but the problem is more layered than just shadow AI. It's not just AI visibility, but understanding how AI is connecting to data and identities. Let me explain it further. First question to answer in discovery is what AI and data exists in my enterprise and where is it being used? We discover all kinds of AI assets, agents, MCPs, agentic applications deployed everywhere, including public cloud, customer data centers and SaaS applications.

One important example of that discovery is our Zscaler endpoint AI asset inventory where we discover identity apps, local AI models, AI development environments that your developers are using, local AI models and AI that your employees are using within the browser through extensions and what skills they are using. This sets the foundation and bring full visibility of not just AI, but the lineage of data and identity within AI, which is really differentiated compared to siloed visibility that customers are getting through their EDRs, identity tools and cloud security solutions. Once we have discovered the AI and data in organization, it is important to understand who has access to that data and what are they doing with it. This is where our acquisition of Symmetry comes into the picture. It allows us to connect with various AI, data and identity tools that organizations are using and build a full relationship or the lineage.

And that is a very hard problem to solve for a lot of large complex organizations. One of the key differentiators here is that we can connect to hundreds of tools in the organization with the connectors that we have built and also the data that we discover in line through our Zero Trust Exchange and what we discover on the endpoint, we bring it together. With this, we can provide the granular insights into the usage of these applications and how they connect to the enterprise data and understand the overall risk. Now risk is broadly defined in 2 broad categories here, risk tied to data and the risk that is tied to your AI. For data, we provide full visibility into risky behavior associated with the enterprise data, including sprawl of data, which is you could have a lot of duplicate data in your environments. That could be very expensive. Sensitive data exposed to third parties or with the elevated privileges.

And most importantly, we classify data everywhere, whether it's on the endpoint, whether it's in line, sitting in your SaaS services or public cloud and figure out the risk associated in these very complex environments. This is what we do with Zscaler DSPM or data security posture management. Now let's look at the risk that is tied to AI, and that includes -- how risky is AI deployments in the public cloud, understand the risk tied to AI models and risk tied to the configuration and posture of these models. And finally, help organizations build a full AI bill of material to understand the supply chain. Your AI applications are not living on their own.

They connect to software packages, they connect to data stores and building that full bill of material is something that we give organizations with the full risk, and that is what Zscaler AI SPM, or AI Security Posture Management does. And most importantly, we remediate the risk with AI agents and integrate with enterprise workflows. Enterprises are really struggling in this area as we interact more and more with them. They end up buying multiple tools and need to manually connect the dots. And connecting the dots is where our acquisition of Symmetry has done a really good job. We not just connect the dots in these complex environments, but we also offer agentic remediation.

Now let's move to the second area of our Security for AI and Data, which is connecting -- which is securing all the AI interactions. As you saw during Adam and Jay's presentation, our Zero Trust Exchange secures every enterprise interaction, including for the AI traffic that might be associated to agents that you are currently deploying or building. So think about a lot of our customers are deploying Claude Code. Many of you use Claude Cowork here or Codex as well as autonomous agents and AI assistants that are running on your computer or agents that run in the public cloud. For agentic traffic coming from employee laptops, enforcement happens on the endpoint from Zscaler with Zscaler endpoint AI security.

And for the AI agents that are deployed everywhere else, we secured with our Zscaler AI Gateway. Before we proceed further, now everyone in the industry is talking about intent and context. And your LinkedIn feeds, you must have seen what is the context-based engine, what is the intent engine. So let's talk a bit about that. When we secure users today, context is typically who is the user, what is their identity? Where do they come from? A corporate network, what devices do they use, what applications they are trying to go to? What is the user's risk profile or risk posture? Is it a risky user because they've done a lot of bad links. And once we determine all these actions, our Zero Trust Exchange grants the access to the user. And this context layer is golden, and Zscaler has one of the richest context layer that we have built for user over the years.

When you think about the agents, the same context layer is now extended to AI agents. Agents also have their own unique context. such as agent risk, everyone is worried about your agents might burn too many tokens. So cost management, token usage is important, but agents also need to be governed based on the intent. And intent is essentially the task, which was -- with which the agent was set up. For example, a customer support agent for a bank should not be giving stock advice. And that task behavior is very important to understand the intent. It is also important to understand how these agents are executing these tasks. They should not use very risky skills. They are now marketplace from where you can get these skills and use them to complete your task. The skill file, for example, can sell agents to steer traffic to an attack and server. And this is where Zscaler differentiation comes into picture because we are the only gateway in the industry today that supports context and intent-based policy to secure the agents.

Now AI agents could be deployed everywhere, most commonly on user devices, but also in public clouds such as AWS Bedrock or Azure Agent 365 or new cloud platforms, which have edge compute. Our customers can send traffic from agents deployed anywhere through our AI gateway in Zero Trust Exchange. And this is where we integrate with agent identity platforms because identity is foundational for Zero Trust and start building that context layer. We just don't secure agents based on what they are, but also understand cost-based routing or more advanced policies that come with it. Our AI Access graph gives us a unique and differentiated bottom-up context layer. What means that we know what data agents are using, what identities they have, and that context there is unmatched in the industry today. So think about this architecture, what we are doing here, we are bringing the identity, we are bringing the context and intent as part of the policy evaluation.

And if I have to summarize the 3 key points on why this approach really stands out for our customers, we not just secure agents going to Internet, but also understand more advanced protocols like MCP deployments. MCPs are de facto in many enterprises to standardize communication, and we have an MCP gateway built here as well. We are also the only gateway to support both deep context and intent, as I explained. And most importantly, built with our DNA of building large-scale gateways, ZIA, ZPA in the last decade to deploy these gateways to handle billions of entities that are going to come at a very large scale compared to many agentic gateways that we see that are deployed only within a specific project or do not scale at that level.

Now let's talk about the one immediate security problem that I'm running into while talking to security teams every single day. How to securely deploy agentic apps such as Claude, Codex or one of many AI assistants securely within the enterprise. Traditionally, attacks are focused on user devices such as laptops, purely on the operating system, file systems and software that employees are using. And this is where we partner with every EDR vendor because they are really good at securing that attack vector. But AI is introducing new attacks targeting users that are primarily focused on the apps that your employees are using, AI assistant they are using and the AI they use within their browser.

For example, employees are downloading seemingly benign files that look clean when they are downloaded initially. But these files can reassemble later and become malicious using what is called a last mile reassembly attack that EDRs can't easily find. What EDRs also do not secure is data exfiltration on the endpoint, whether that is transmitted locally using a USB stick or using a local MCP server that a developer has installed on their computer. This is where our endpoint AI security product comes into the picture as AI apps usually get full access to data on the user device and can transform that data via prompts and transfer it to unauthorized destination. We provide 3 fundamental set of capabilities here. AI usage control to provide control on usage of sanctioned AI and detecting non-sanctioned or shadow AI on corporate devices and do granular controls with these apps.

For example, we do not allow data transfer between a corporate version of ChatGPT and a personal ChatGPT in the same browser. Cyber protection to understand malicious skill usage, use of risky AI models on the endpoint. You know that you can actually run a small local model embedded inside your browser or on the device now. And these could be used to create sophisticated phishing attacks through prompt injection in multiple turns. And these are the kind of things that EDF can't find. And finally, data protection, Zscaler has the endpoint DLP product that discovers, classifies, monitors data everywhere and allows you to build consistent DLP policies around your data including what goes inside the prompts and access or data that is going out through browser plug-ins. This is a very refreshing approach for enterprises that we work with -- the security teams we work with because they do not need to deploy multiple agents and get fully integrated AI and data security solution, which is very complementary to their EDRs.

Data also needs to be secured in line or when data is in motion. And there is -- that is where we had a significant advantage as we secure data with Zscaler DLP across all channels where data can exfiltrate. Data, as Jay explained, in ransomware attacks, either is going to Internet to attacker servers or to unsanctioned applications. Data can exfiltrate over e-mail or prompts within these applications or embedded chatbots that you're using. For example, if you use a food ordering app, it has a GPT-powered chat bot in it. Underlying intent of ransomware attacks is always to steer data. And if you have deployed Zscaler DLP to protect those channels, we prevent this from happening.

So what I do want to highlight here is the fact that because of our Zero Trust Exchange architecture, we have a unique advantage. We have a proven proxy-based architecture to intercept and secure any data leaking to the Internet. We provide a uniform policy layer for data security across all channels. And all of this is delivered through a single Zscaler client that is running on more than 55 million devices today, and customers don't need to deploy a new agent, which is really very important point for organizations as they want to simplify what goes on their corporate endpoints. Let's look at an example of a large global financial asset management firm that leaned in with our data security and how that created an upsell for our AI security solution.

The challenge this organization was facing is lack of visibility into data in their public cloud environments and limited classification of data sitting there. Using different data security products to secure data in public cloud and different for on-prem and something different from endpoint, they lack the full picture of where this data and what's happening with that data. With Zscaler DSPM solution, they up-level their data classification and got a much broader coverage of data asset stores that they have deployed. As they connected more data stores, they understood the AI sprawl and how AI is connecting to their data. And that led to the evaluation of Zscaler for AI security offerings. And within 1 quarter, we upsold them Zscaler AI security. The benefit from customer perspective, one of the main benefit was discovering petabytes of redundant data, which they removed and had a lot of cost savings around it.

They also understood the data access across their organization with upgraded data classification and most importantly, single solution to secure data everywhere with pre-supported connector for every data store that they use within their organization. This also led to $5 million worth of expansion for Zscaler in this account, and our DSPM outperformed the stand-alone DSPM they were using, which they replaced with this. Now let's move to the third pillar, which is helping organizations protect first-party AI apps, MCPs, models and data stores they are using within their organizations. To protect AI systems, we focus on securing full life cycle of AI within the organization. So starting from when you are building these applications to when you start deploying these in production using our Zscaler AI Red Teaming product, as I explained that came through acquisition of SPLX.

And then once you deploy them in production, we secure these apps with our Zscaler AI guard solution. Today, we are the only vendor in the market that is a fully integrated AI Red Teaming and a guardrail solution that takes input from the red teaming and enforces automated guardrail policies to secure runtime behavior of these applications. Now let's take a look at Red Teaming for a moment. We provide -- in this solution, we provide a catalog of dozens of supported agent platforms or model providers, hundreds of model providers supported and thousands of MCP servers on Internet are already prebuilt and supported with it. Next, we have a library of pre-configured test profiles around safety, security, hallucination of these models. And once you configure those tests, we can simulate 5,000 attacks like how attackers would attack your applications and test it with multimodality of like throwing text, images, files to see how your applications break.

This helps enterprises find vulnerabilities and risk in their apps before they go live in production. But with our model benchmarking and baselining for frontier labs, open weight models, we help organizations compare and pick the best models. One important thing to remember is that the entropy or the change in AI applications is usually very high. Every time a new model version comes or a new software library is released, overall risk posture of these applications change. This is why we support automated on-demand Red Teaming. And finally, we are the only AI Red Teaming solution that provides an integrated and a stand-alone prompt hardening service, making sure your apps are that you're building like chatbots, when they interact with the model, they have a very consistent hardened prompts they are creating. If prompts are not harder, we offer the remediation as well.

Next, when these apps move into production, such as everyone is building an employee-facing chatbot today, which almost every enterprise has built with a small or a large language model they have deployed or a front-end model that they deployed within their organization. And this is where we basically analyze the risk requests going into these chatbots and the response that come from the applications through their models and a true intent-based inference or a runtime AI security policy engine that secures every prompt and reviews every prompt as it is getting to these applications. This strategy of securing prompts across heterogeneous environments, such as multi-cloud systems. So a lot of our customers basically are running their or consuming their model from one hyperscaler, their application is running on different hyperscaler. And one good moat that we have built here is that our customers, like how did -- Adam talked about Zero Trust Cloud to provide same security across all hyperscalers, we provide consistent guardrails across any model provider, any hyperscaler.

And since we anticipated the need for such intent-based policy a few years ago, we had a head start in this area when we introduced AI DoT 2 years ago and have started creating a catalog of predefined detectors across different industry verticals, and different use cases that are trained with use case-specific data sets and small language models that we run on GPU-based infrastructure in our cloud, so customers can easily deploy it with 3 clicks. We have also extended the guardrails of this AI Guard to prompts that are going through our Zscaler Internet Access for 3,000-plus application. And finally, here, we are able to enforce policies based on the token usage by these application, we can tell you which users are using how many tokens and we can do cost control on that as well.

As we bring this together, right, to summarize, what we have covered so far across 3 pillars and how it comes together as a single integrated solution for our customer is that it starts with the discovery layer. Of every AI asset, including on the endpoint, models, applications, MCP tools and understanding its deep relationship with data and identities to create an AI access graph that allows us to build very comprehensive DSPM and AI security posture management solutions. This also builds a highly differentiated context layer that enforces policies to secure every interaction with our AI gateway and of our agents deployed everywhere, while we extend AI security to user devices with discovery with our AI endpoint security. And by the way, with AI endpoint security, we have also embedded Mythos 5.1 model in early access to find more sophisticated attack parts on the endpoint as well.

And then protecting AI apps through build, deploy and run time life cycle with Red Teaming and AI DoT. Let's look at another example or a case study of a large global system integrator with more than 400,000 employees who lean into our entire AI security portfolio that I just talked about. This organization had a very open access to Gen AI applications for a couple of years, then got audited for very high risk due to data exfiltration risks. And as a result, access to all AI application was revoked within this organization. Business started losing competitive edge, as you would know, in today's world. Customer who was already a Zscaler Zero Trust for users customer started enabling sanctioned AI applications for employees with GenAI visibility and AI guard for users to secure traffic.

They also started mandating AI Red Teaming for first-party apps they are building and also started deploying agentic cowork application for their employees using our endpoint AI security. This customer saw the value of consolidated AI platform, which already integrates with investments in data security, ELP and a full flex deal for our entire AI security portfolio. That gives them access to the entire portfolio that I talked about and everything else we are building in this area. Broader access to -- from a benefits perspective, broader access to a catalog of AI apps and being able to do on-demand Red Teaming for first-party apps while keeping the compliance with European AI -- EU AI Act or NIST, AI Risk Management was a big differentiator. And this customer did a $5 million-plus security for AI deal with us in fiscal Q4, raising the overall ARR for this account to $18 million.

Now as we wrap up the session, I want to summarize what really excites me about this massive opportunity and why we are bringing AI and data security together, not just for -- by the way, for us, but also for our customers in terms of the outcomes that they're driving with us. So our leadership in data security creates a strong demand for AI security as AI security outcomes are closely tied to data security and data problems. Pace of AI innovation is moving at lightning speed. Customers are looking to buy consolidated platforms versus individual point products to secure AI usage. And we believe the Zero Trust architecture is fundamentally built to handle the scale of AI agents, and this is where we have an undue advantage. And finally, both Security for AI and Data Security Solutions allow customers to deploy these products with no dependence on rest of the Zero Trust SASE portfolio. It can be onboarded and sold on its own. And with that, I would like to wrap this presentation. Thank you for paying attention. And next, we will move to the customer panel.

You have heard this morning us talking about our architectural advantages, power of Zero Trust Exchange and how our solutions are helping our customers navigate a rapidly evolving threat landscape. But ultimately, the most important perspective is coming from our customers. And this is why I'm glad to be joined by 3 of our esteemed customers here AkzoNobel. They are a leader in global paints and coatings with a presence in 150 countries, and ideally know this customer because they became a customer the year I joined Zscaler in 2012. So they're a long-term Zscaler customer. Marsh McLennan, the world's largest insurance broker and a Zscaler customer since 2019; and Merck, a leading global pharmaceutical company, our customer since 2022.

We were intentional about bringing together customers from different industries to illustrate what the breadth of all the challenges that enterprises are facing and increasingly universal need for modern security architecture across different verticals. Our panelists will share more about the challenges they are facing, how they prioritize their security challenges, how their priorities are changing in the AI era and what they're looking for AI security solutions. So let me introduce our panelists. Please welcome Kurt De Ruwe, who is the CIO of AkzoNobel. Kurt, please join us here. Jon Eastern, who is Global Head of Networks and Security at Marsh & McLennan. And finally, joining us virtually is Bob Stasio, who is the Deputy CISO at Merck.

Thank you, everyone, for being here. I will kick it off with a quick question, and then we will open up the floor for the audience in the last 10 minutes as well such as challenges that your company faced when you first deployed Zscaler and what your Zscaler journey has looked since then?

Yes. I think, first of all, it was the protecting of the Internet. It was at the time that we still had MPLS networks. I think if you look at the journey, all the products that were at the bottom that Adam showed, I think we have deployed them. So visibility with ZDX, we have ZPA. We have RISC 360. And I think if you look at it more recently, like 2 years ago, we wanted to go to open Internet everywhere. So that's 995 locations. So we do that on the one hand, with ZPA, on the other hand, with the branch connector. So it's really, I think, Zscaler is fully underpinning or giving us the technology to do what we would like to do.

Interesting. And coming to you, John, you've been a customer for a while as well, and our relationship has been evolving. I remember especially interesting time during COVID on ZPA deployment. Can you share some of your initial thoughts on how you deployed Zscaler initially and how that has been evolving?

Sure. Yes. So when we started with Zscaler, we were -- I think like many other companies, we had a centralized Internet. Everything went through data centers in a couple of different regions, legacy proxies. We wanted to digitize our workforce and Zscaler was really what we thought was the best way for us to secure that Internet traffic in particular and enable that business to use those more modern tools. ZPA, we began testing initially. We had a specific business use case for it and then COVID accelerated that plan pretty quickly. But really to the credit, and we deployed it over a matter of days to our workforce of, I think, probably 75,000 to 80,000 at the time. We're 100,000 now. Everybody using that as our primary tool for remote access and even in the branch office in our Zero Trust model, I'll talk about a little more later. But yes, our relationship has continued to grow as your product portfolio has grown successfully.

So key point is that you started with focus on protecting users and the use cases have been evolving since then. Bob, I'm going to come to you. You have been a full platform adopter. What advantages you believe you've got from leveraging the full breadth of our platform and how that has been evolving for you in the last 4 to 5 years?

Yes, sure. Thanks for having me. Yes. I mean the way [Technical Difficulty]

You are looking at the whole spectrum of Zscaler offerings. You're also a Z-Flex customer, right? So can you talk a bit about as you looked at the entire landscape of what Zscaler is doing, and we talk regularly on what is on the truck coming soon. How did you move from product-based licensing to Z-Flex? And what was the thought process behind it?

[Technical Difficulty] Pull in another capability to test out without having to go back and do specific product-based licensing. So we've been able to check out things like the branch connector and even some of the more, I guess, newer offerings, advanced offerings like the Zscaler Cellular in some of our manufacturing areas. So that's been very helpful for us as we've deployed these capabilities across the enterprise.

That's amazing to hear that we have customers who have started adopting deploying cellular technology that we had the same on actually production environment. We are seeing a significant uptick in that now. Coming to Kurt, you've been an at-scale customer for our Zero Trust branch offering. You started pretty early when we had the branch connector. You also have some very complex OT environments and securing OT comes with a lot of legacy infrastructure as well. Can you talk a bit about when you thought about bringing Zero Trust to the branches, and I'm pretty sure you had SD-WAN. What made you think about adopting Zero Trust branch? And how do you see that as fundamentally different from SD-WAN?

So basically, we had the idea, Internet everywhere with the aim of taking out all firewalls everywhere and also getting rid of SD-WAN. Now if you look at the offices, if you look at warehouses, that's fairly simple, if I can say. If you take them into the factories, it's more difficult because you have the PLCs, you have SCADA, you have devices on which you cannot implement agents. So that's why we looked at the branch connector. And it's actually -- the 4th of October was our first deployment last year, not 2 days ago. So we're currently at 69 sites -- factory sites that have been deployed out of 110. And it's -- so the thing is that we have a mix of everything, any equipment you could buy in the last 30 years we have. We have fairly old, we have fairly new. But basically, the intention is to implement the branch connector and the micro segmentation.

So if you look at it, I think in addition to that, also what said is that with the aim of going to Internet, open Internet, it was also aiming at a significant cost reduction and also the Zscaler SIMs. So rather than replacing all the network equipment, for instance, in a warehouse, we put the SIMs into the scanner, so we don't have to do the big investments. And maybe one anecdote. When we did Como, that was the first site. We organized a price setting for our security team. So I said the first person that can break in and they were allowed to use any tool would get a nice price. And I think 2 days later, they came back and said, well, we can't find it. So I think that's -- it was also one of the statements. If you can't see it, you can't hack it. But for me, that was really an important proof point that this is the right way to go.

That's amazing. Actually have a lot of customers do that. As part of testing, they bring rogue access points in their shop floors, plug it in and say, what can I find the technology the way we have built, you don't find anything else with that. Now Jon, you have unique advantage of network and security both, right, in the firm. So one thing that everyone wants to know is like -- how do you realize the advantage of using a Zero Trust architecture? Sometimes security and network look other way. So how did you bring that convergence of Zero Trust for architecture to protect your users everywhere?

Well, I think over the past 5 years, network and security have really converged, and that's why we made the organizational decision to combine them. But in today's business world, we need to really support our business to be able to execute at almost machine speed. AI has only accelerated that. So our ability to -- using Zero Trust we reduced the elements of friction for our users, giving them better access, faster access while simultaneously improving the security model. So it's a nice convergence that they don't often see. We find our users are preferring the new model, leveraging Zero Trust. They are able to do more and we have more visibility and able to secure it more. And those tools have really enabled that ZPA has become the foundation of our Zero Trust model, providing that access broker for all applications, making sure that all of our users can access the applications they need to and none of the access -- don't have access to the applications they don't.

Yes. And I think experience is very important with security. And I think what ZDX does in this space is giving you visibility of user experience with a Zero Trust architecture also unlocks a lot of bottlenecks that we traditionally have seen with network.

Absolutely.

So let's switch gear towards AI, right? I talked a lot about AI and how awesome our AI security is. But I want to start with these speakers, Bob, coming to you, you guys have a big directive to deploy -- use AI for innovations. And can you talk us, first of all, key use cases where Merck as a company is leaning in with AI? And what security challenges it create for you as a security leader as AI gets adopted widely.

I muted. I didn't want you to hear my Teams alerts. Sorry. So yes, we made a public announcement earlier this year about a $1 billion investment into Google Gemini as kind of our main orchestration platform across our enterprise. And we have several, we call lighthouses or kind of lines of effort across the enterprise to help our organization. I mean at a high level, what we're trying to do is typically, a drug takes about 10 years of development and of all the drugs you look at of all the molecules you start to investigate, maybe 1% of them will go to market. We're trying to get that down to 5 years and 10%, right? So it's a really interesting multiplication there of effort, and AI is going to really help us drive that forward. So there's various efforts to help the research element of Merck, even the manufacturing, even some of our global support functions to include IT and finance, be more efficient with what we have and roll out -- excuse me, and roll out more drugs in our pipeline over the next 5 to 10 years.

So AI is a big part of that. And obviously, this business is focused on it. So we have to be focused on coming behind and making sure it's secure. So what I'd like to talk about is the AI governance, safety and security governance life cycle. So we're focused on how we're intaking various AI applications, agents and models. We're categorizing them by risk. We're applying appropriate controls. We're monitoring those controls, and then we have compliance and auditing against them. So we're using some of the Zscaler AI products to help in that journey, and we've started rolling them out actually last quarter.

Yes. That's great. Jon, I'm pretty sure you are looking at AI very closely in the firm as well. Some of the challenges that you are seeing, not just from security, but from also how you connect your networks traditionally to these AI ecosystems.

Yes. And so from a business perspective, we have kind of 3 pillars of strategy for AI. One is -- the first is individual productivity using internal chatbots like similar to ChatGPT, working at business process optimization, how can we take our existing processes and use AI to improve efficiency and optimize there. And then more business process transformation. How do we transform our business to operate more effectively and win in the AI-enabled world. So we've had to -- so with all those challenges, you need to secure the environment using AI and you need to protect the environment from AI. And that's become really the critical challenge that we're working on and working closely with partners like Zscaler to do that with the portfolio.

And you have been giving great feedback now with guardrails and detectors as the use cases requirements in each organization keep evolving. So really appreciate the feedback that we got from you. Your world is very different, Kurt, when it comes to agents and security, and we were at dinner last night, you actually, we were surprised to see how vast is your footprint of AI from not just employees, but also in your plants and manufacturing in other places as you're thinking about AI usage. So how are you thinking about security evolution in order to protect AI in these complex environments?

So I think currently, we have 8 AI towers where a lot of the teams we offer internally. Like all users have access to more than 12 models, including some of the Chinese models, which may sound strange, but we allow within a framework to use it. And on the 26th, basically, we will offer a toolkit for all our users to create their own agents. So it's really the challenge to make sure that the guardrails we have that they're good enough and then basically to see basically what everybody is doing. and to be able to intervene where things get -- might get out of control.

Okay. And we know that you have recently started looking at our AI security portfolio, and we are going to take a deeper look at it.

Now we are piloting solutions. So...

All right. In the interest of time, last question for you, Bob. We recently launched our SecOps offering, and you have been a Red Canary customer prior to acquisition as well, and you have been a strong voice in how we should be evolving the acquisition of Red Canary into the Agentic SecOps that we have built. Can you share some insights what you're looking in terms of how this should come together?

Yes, I'm actually a 3-time Red Canary customer. So my original view of the Managed Defense MDR solution in Red Canary is to really push all of your, what I call high fidelity alerts into the platform, and it can really triage it at scale. So those types of alerts would be coming from your EDR or maybe things like Wiz or GuardDuty. Low fidelity alerts would be things coming from Palo Alto or even in some cases, like the actual Zscaler alerts like ZIA alerts coming through because it's just such high volume, most of them are either false positives or just indications of blocking. So I initially tried to focus the Red Canary solution on the high fidelity alerts, which we had about 20 agents in the platform that are running over all the data and really automating everything we did, and it really helped us out lower our mean time to remediate significantly and lower the number of incidents we're seeing per day or investigating per day and allowed us to kind of move up the maturity curve.

And that was kind of Stage 1 or Phase 1, which worked really well. And now what I'm really excited about for Phase 2 is that kind of lower fidelity data, how can we use agents to go across that data and do more kind of like hunting. So the first one -- the first case was more of like triage and maybe automating the initial stages of an investigation. In the second phase, it's more like true hunting, looking for anomalies, pulling threads. And I think AI is very good for that, kind of recognizing patterns. And we're kind of experimenting with that now as a design partner with Zscaler to roll that out. I think that's going to make a huge difference and really kind of complete a full AI-driven SOC.

Okay. Thanks for as well. Now we will open the podium for the questions from the audience. If you could please raise your hand, we have mic runners in the room who will come to you. Please state your name, your firm's name and ask your question to the panelist. Let's start over here in the front.

Mike Cikos with Needham. We're all hearing the presentations you guys are here listening to this on AI security. And I'm just trying to get a sense, if you're thinking about your cyber budget today, let's say, it's $100, right? And you're spending make up a number, $10 on Zscaler 2 years ago. What is that percentage of spend today? How does that change in the next 2 or 3 years just given incremental AI security budget that you guys will be spending on these solutions?

I think this question was not addressed to anyone, so I'll take the liberty for maybe starting with you, Kurt.

I think to give percentages is difficult. The only thing I can say is that almost every new thing that comes out, we look at and we adopt.

[indiscernible]

I think it's mainly additional because the good thing about Zscaler is they think about the problems you will have that you don't know yet that you will have them. So by the time you encounter the issues, there's a ready solution there. So in several cases, we did not even do an RFP like on ZDX, we just basically -- we tested it. We like it. We agreed on a price and we deployed it. The whole -- I think what is very important with Zscaler is the concept on -- if you have the client, I think it's also what you said, deployments happen within days of new capabilities. Bob, anything you want to add to this?

Well, I guess we're in a slightly different boat because we have sort of like an ELA with Zscaler. So we kind of have nearly all of the portfolio or access to it. Obviously, there are some separate items that we buy a cost. But I'll say it's -- I don't want to give percentages, but I'll say it's one of our top security vendors that we have on spend. Spending in security is obviously a little difficult to calculate because we spend a lot on Microsoft and security, but that's part of our larger Microsoft budget. But for our own budget within like our security group, it's one of our top items.

Let's move to the next question in the front here.

Absolutely. Saket Kalia at Barclays. Thanks so much for hosting this panel. Super helpful. Maybe this is a question for everybody, if possible. So as your usage of agents across the business, across your respective businesses grows, how do you think about your usage of current Zscaler products like a ZIA, ZPA? And which other Zscaler products maybe become more relevant as that adoption grows? Does it make sense?

Jon?

Yes. So what I would say is in the AI world, Obviously, things change, but we don't see a strong well-architected layered security model is still the best defense against these -- against the security environment. So nothing changes for us from that core set of tools that we purchased from Zscaler or anyone. There are absolutely AI-specific use cases that we need to address and Zscaler is building a really interesting portfolio that we're evaluating. But really, our focus is on AI speeds everything up, especially in the attack space. But that core is still really important and something we -- I'd say we're still very focused on and even doubling down on.

Maybe one example is we want to go to fully ticketless, which basically we need to have all the inputs and the signals, so what you see is that Zscaler is also AI enabling, if I can call them the older products like ZDX. So we have an intelligent hub and then ZDX feeds in basically what is going on. So that also basically allows us to do more in a fully automated way. And I think as security becomes ever more important, it just -- is the whole suite that adds the value.

That's a very good point because what I've talked about was security for AI, but we are also doing AI for security in every core product. And that also is leading to a lot of exception. Like a good example is deception product where we actually have LLM decoys -- decoys for LLM. So that becomes a natural expansion of Zscaler. Let's move on to this side. Right in the front here?

It's Keith Bachman from Bank of Montreal. I wanted to ask a 2-parter. Where do you think you are today in your AI journey in terms of AI adoption versus where you might be 12 to 18 months from now? And the corollary part of the question is, how do you think about not necessarily the companies per se, but just the incremental security that you need to adopt to enable further AI penetration. For instance, a lot of vendors talk about governance solution, cloud security agent identities. How do you think about just where you need to adopt technologies, buckets of spend, if you will, in security? And how does that compare and contrast to your aspirations for AI adoption over the next 12 to 18 months?

Bob, do you want to take?

Yes, I can jump in. And your question is -- you're asking where are we with AI adoption like at the enterprise or within the security organization?

Within the enterprise broadly.

Okay. Yes. And here, we have a pretty accelerating model on AI adoption. So the core folks in our -- we call them lighthouses, too, within our lighthouses like in the research labs and MMD are fully adopted in those use cases. But across the enterprise with like business process assistance, the copilots, things like that, we're moving towards the higher percentages there. It's accelerating every month. So we're keeping an eye on that. But those agents are pretty controlled in orchestration platforms. So like for example, like the Gemini that we use is kind of this orchestration layer that automatically has controls and harnesses across all the agents developed in there.

So we feel that they're pretty safe rolling out there. We're trying to drive users to adopt there instead of bringing in the new type of agent or model. So that's good. And then your question about like where do you see the spending going and especially around like security governance. So I think every organization is struggling with this. Every vendor I talk to has a different approach on security governance. I don't think it's figured out yet. And what we're -- kind of the first thing we want to do or we're doing is this kind of like a blocking and tackling mindset. So just because the AI is rolling out very quickly, it's in use, we need to be monitoring it, controlling it and blocking and tackling issues that pop up. So we've established a couple of different vendors to help us with that and a bit of kind of tiger teams to stand up and do that. But that's kind of a temporary solution.

It's a process that we had to develop because it was needed. And we're having a study going on right now in our organization to identify how we're going to do the governance process over time, build a larger orchestration platform that's going to work across legal, governance, ethics, compliance and IT and engineering, which is pretty complicated. So that's kind of where I think the money is going to -- the funding will go in the next couple of years is towards a more robust solution that can take these atomic level feeds from all these areas to monitor what's going on, categorize it and satisfied stakeholders that I mentioned, which is pretty challenging. And we have not really seen anybody come across and do that yet. Zscaler is certainly a part of the solution, right? But it's not the whole solution because there's a lot more that has to be involved in it.

Yes. I think the key point that always making is evolving landscape. And one of the things that we are able to do is provide flexibility for customers to see what is relevant and needed for them, and they can switch to what is needed.

Maybe just spending money on security for AI is not a choice. You have to do it. Spending money on agents, there you have the choice depending on the consumption.

All right. Let's go in the middle somewhere over here.

Brian Essex from JPMorgan. I had a question for Kurt. Within your manufacturing facilities or shop floor, I'd love to know how you're addressing the risk to your operational technology within those facilities. I think you mentioned you might have some stuff running on SCADA. So how are you addressing the architecture and risk posture within your manufacturing facilities? And how does Zscaler help you do that?

Yes. So previously, the way we run our factories is like islands connected from the main network. But once the hacker is in, they can go everywhere. So I think, first of all, with the branch connector, we try to make the site invisible. And then in case there's still a breach with a micro segmentation with limited to just a number of systems. With regards to the PLCs, Zscaler has solutions for that, so for the horizontal flow. And I think it has been a challenge, to be honest, because a lot of the technologies that we have are different. So it was not you have done one, you can now do the 100 other ones. So every site, we need to learn. But basically, it's -- the impact that Zscaler brings is totally reducing the attack service. And when something happens, we have the visibility.

Are you adopting AI within those facilities? And how are you addressing the potential risk if you are, in fact, doing that?

So I think AI is the next step. It's initially in putting in place the protection and then you have the triggers and then those triggers basically are being followed up. I think more and more AI will be used there to then just take those triggers that are really relevant.

I think this is a great point to like wrap the session, and we have a great point where we actually have had a wide variety of topics that we discovered. So I want to thank all of our panelists and for joining us here today and for the insightful discussion with me and all of you. Before we break, I want to take a minute to reflect on a few key themes that have emerged since in this conversation and what we have been discussing since the morning and based on discussion that I've been having with customers.

First, these conversations continue to reinforce our position as a strategic partner to our customers. What often begins as a single use case, for example, to protect users, then evolves because of the architectural advantages into expansion of Zero Trust Exchange beyond users to workloads and branches and expanding our platform deployed with the customers. Second, our customers are increasingly relying on Zscaler to solve complex problems. The Flex programs gives them that flexibility. They can co-innovate with us. They can be design partners and solve security challenges and go hand-in-hand with us on some of these emerging priorities.

And third, AI is a big opportunity, and it is also a challenge from an architecture perspective, and we are looking to help our customers navigate both. The need to securely enable AI while also making sure that threats are -- you're protected from threats is a critical requirement. So against this backdrop, the opportunity is significant, and we believe Zscaler is uniquely positioned to capture it, helping our customers succeed in this very complex world. With that, thank you again, Kurt, Jon, Bob for joining us here. And I will now welcome back Kim to the stage. Thank you.

Okay. Thank you very much. And now we're heading into a short break, which you're probably very relieved about. We'll take about a 10-minute break. But before we do, I just want to put the demo slide up once again. So right outside the door, we have 4 demos. So a lot of the solutions that we've talked about today from our Zero Trust SASE offering to Agentic SecOps, which Adam talked about both of those, the Zero Trust Exchange for agents, there's a familiar face there. That's also what Dhawal talked about and data security solutions, which Dhawal talked about as well. The person who's hosting that is actually the CEO of Symmetry -- the former CEO of Symmetry, now Zscaler employee. So enjoy those, and we'll see in 10 minutes. Thank you.

[Break]

Please welcome back to the stage, Kim Watkins.

We are ready to get started again. So come on in, take your seat. Before I introduce our next speaker, I wanted to mention that we'll be sending out a post-event survey after today's program. We would really love your feedback. So please, please fill it out. And our goal is to make sure these types of days are most valuable for you. I also wanted to give you a brief reminder. Today's presentations contain forward-looking statements within the meaning of the safe harbor provisions of the federal securities laws, including statements regarding our future financial performance, business strategy and market opportunities. And with that, it's my pleasure to introduce Ross Tackett, Zscaler's Chief Revenue Officer. Ross?

All right. Thank you, Kim. Welcome back from break. I'm going to go over our go-to-market strategy for the next 4 hours. So I hope you buckle in. I kid. I am going to go over our go-to-market strategy for about the next 20 minutes. But before I do, I wanted to provide some background on myself and introduce myself to you all. I've been at Zscaler for about 3 years now. The first 2.5 of that, I was successfully running our largest geography, the Americas. And in the last roughly 6 months, I stepped into the worldwide sales leader role, where I was responsible for delivering on our global sales number. As part of that role, I've been heavily engaged in the annual planning process for both our operational and financial plans.

We have a strong strategy, a strong plan in place and the entire team is focused on executing. I am very confident this will be a seamless transition. There are some key messages that we want to make sure we get across in this session. The big one, as Adam and Dhawal and Jay have mentioned, is Zscaler was and is built for this AI moment. We have the right platform, the right team, it's the right time right now to be successful. We've built this durable go-to-market engine, and we'll talk about how we're leveraging that go-to-market engine to capitalize on these AI tailwinds. I'll begin with a few key reasons we are well positioned to capitalize on this moment. First, scale and global presence. Our customers are global, their security challenges are global, and we have local geographic coverage all over the world to provide the best customer experience possible.

We serve customers in more than 185 countries globally with 25 countries generating over $10 million in ARR and 7 countries exceeding $100 million in ARR. This worldwide reach is a differentiator for us. And our international scale and strength is evident with greater than 45% of our ARR coming from international markets. Next, we serve an incredibly diverse customer base spanning across all industries and verticals. Overall, we serve over 50% of the Fortune 500 and over 40% of Global 2000 customers and accounts. This broad presence demonstrates the universal relevance of our solution to enterprises globally. You can see a few examples on this slide of verticals like hospitality, technology and retail, where we work with a majority of the top companies in those sectors.

This broad coverage has a strong network effect. CISOs and CIOs often buy based on advice from other CISOs and CIOs. So this places Zscaler in a unique position for continued growth. Now I want to talk about how we go to market by segment, how we're organized internally. Our major segment at the top of the pyramid generally includes accounts with 40,000-plus employees or a very large propensity to buy from Zscaler. These are covered by sales executives that have just a very few accounts so that they can focus on a high-touch sales-led approach, often in conjunction with our GSI partners, and these partners are in every one of the largest global accounts. Our middle enterprise segment covers 5,000 to 40,000 employee accounts, and these are jointly led by Zscaler and a partner, either a GSI or a national VAR. We see this part of our business as an area where we can accelerate new logos, and so we've invested in creating more hunter territories to drive this growth.

And finally, our commercial and SMB space, these are businesses and organizations under 5,000 employees. This is increasingly shifting towards a partner-led motion with simplified bundles. This segment is a high velocity, new logo generating machine that is bringing on new customers who will one day be our next enterprise accounts. But regardless of segment, building trust, building trusted relationships and having executive alignment is an important part of our go-to-market approach. This allows us to be a consultative partner as part of a customer-centric strategy. Put simply, we focus on understanding the problems they are trying to solve to help them transform their security architecture and address every security challenge. Now I want to double-click on our vertical strategy because this is a very important part of our growth plan.

We have established dedicated teams for the last few years focused on certain verticals, such as U.S. public sector and U.S. health care provider. This year, we launched a U.S. financial services vertical as well. All these are critical verticals that require additional coverage and expertise. For example, in our federal business, our FedRAMP High and IL5 accreditations and ability to deliver solutions for highly regulated environments are a competitive advantage for us. And we see opportunities to expand our vertical strategy in the future. As AI and the increasingly regulatory environment create tailwinds in the space, we see a path to future growth and opportunities in areas like international public sector and sovereign cloud.

Sovereign cloud is a big focus for us, and we are seeing a lot of demand both in public sectors as well as regulated industries around the world. Okay. Now that I've covered how we structure our go-to-market organization, I'm going to focus the rest of my presentation on the 4 key levers driving our growth. First, platform expansion. The AI security tailwinds are undeniable and tremendous, and we are seeing significant momentum from our new product offerings. Second, we are accelerating our new logo engine, and the team is focused on landing new customers with these expanded offerings. Third, and crucially, we are scaling through partner leverage, working closer than ever with our ecosystem to reach more customers wherever they are. And finally, everything we do is underpinned by customer obsession, ensuring relentless focus on our customer success translates into tangible results with platform expansion and new products being purchased.

I'm now going to touch on a little more detail on each of those areas. First, starting with platform expansion. As you've heard, Zscaler is truly a cybersecurity platform. And here, you can see a snapshot of this platform. You've seen it before many times today already. Our 4 product categories are built around a common architecture, as you've heard from others. What we haven't shared is there are many -- there are more than 30-plus distinct products underpinning each of these categories. And that number continues to grow as Adam and Dhawal and their product innovation teams build more products. This is a key differentiator for us versus other companies that claim to be platforms, but are really just a collection of point products that have been acquired. We see substantial demand as our customers continue to invest with us and consolidate point solutions in exchange for a platform they can trust and one where they can realize significant long-term savings.

Today, the majority of our ARR comes from customers who have at least 1 product across 3 or more product categories, demonstrating the breadth of our platform and the value proposition of what we offer. These customers not only have significantly higher ARR, they also have a higher retention rate. And critically, there is still significant upside potential for Zscalers inside of our existing customer base. Today, on average, Customers have purchased only 9 out of the more than 30-plus products we offer. So there's a lot of opportunity to grow. And when you look at the white space across our installed base, the opportunity is significant. As you can see, we conservatively estimate our current TAM, not including our future TAM, but our current TAM of $19 billion.

When you consider our FY '26 ending ARR of $3.8 billion, we have a significant $15 billion worth of white space available to capture. That represents a 5x expansion potential just by cross-selling our existing products to customers who already know us and trust us. Notably, this expansion potential is spread across majors, enterprise and commercial segments with roughly 3/4 of the expansion opportunity among majors and enterprise. But regardless of the customer, the story is the same. Once a customer experiences the power of the Zscaler platform, they want to consume more of it. And as our customers navigate a rapidly evolving threat landscape, we are perfectly positioned to grow alongside them. Let me give you an example of that growth. This is through the lens of a global investment firm. This is a customer that we landed with a $2 million ARR that was Zero Trust SASE for users, and that was in FY '19.

In FY '22, they expanded into data security, and then they added SecOps in FY '24. And recently, last year, they added Zero Trust SASE for cloud and security for AI, growing this ARR to more than $15 million or 7x the initial land. This growth is fueled by deep executive alignment on strategic transformation, which drives the customers' continuous investment as our platform expands. But the most important part of this chart is the trajectory to the far right. And even at $15 million, we still have a substantial upsell opportunity ahead of us as this customer continues to consolidate on our platform.

Next, our new logo engine. We have a significant untapped opportunity remaining in our target market of organizations with more than 1,500 employees. Our current customer base represents just 4,600 of these estimated 20,000-plus enterprises globally. There is over 75% of the market that we have not yet reached. And as we head into FY '27, we have made some changes to ensure the field is best positioned to capture this significant opportunity. For example, as we approach territory design for FY '27, we put a strong emphasis on new logos. We significantly increased the number of 100 territories globally, and we have also focused and incentivized our field for new logos and driving that growth. And finally, we work very closely with our partners to provide further leverage in opening doors for new customers that have been shut for us. And I'll touch on that a little bit in a moment. You've heard it from Adam, you've heard it from Dhawal, you heard it from Jay.

Our platform has expanded. We now have a multitude of solutions to land with new customers with entry points across each product category. Here are just a few examples of use cases where we can help prospective customers address with our growing platform.

I won't go through all of them, but I do want to call out security for AI as it is a huge area of interest in every conversation that we are having. You heard that from our customer panel today. I'll also call out Zero Trust branch. We're winning very large new logo 7-figure deals with ZTB, and they do not have users, great landing spot.

And let me give you an example of just one customer. We recently had a 7-figure win with a Fortune 500 life sciences leader. This deal is a perfect illustration of the power of the Zscaler platform and our security for AI innovations.

It was led by a newly appointed CISO who is a repeat Zscaler customer. You heard that from Jay. This is a very common theme. And we won this customer, thanks to our superior security for AI capabilities.

Our platform gave this customer the specific controls they needed to safely adopt AI without compromising sensitive data, allowing us to displace a legacy firewall-based SASE platform. Next, I'm going to talk about partners.

So all right. Okay. I want to talk about partners. Sorry, technical issue. I got a pointer problem.

Our partner ecosystem is the engine that allows us to scale across the entire market. At the top of the pyramid in our majors and enterprise segments, we are seeing incredible progress with GSIs with greater than 75% year-on-year TCV growth in FY '26. These partners are leading multiyear board-level digital transformation initiatives where security for AI has become a central theme.

Strategic VARs are also -- remain an important partner for us. We work with these partners across all of our segments, especially in our enterprise business. These organizations have local expertise, long-standing relationships, and we recently launched a new incentive program to further penetrate this market. Now moving down market into commercial and SMB, our strategy shifts towards scale and velocity.

Here, we are intentionally leveraging distribution and managed service partners for scale and speed. One example that we announced is our recently expanded partnership with Carahsoft targeting commercial and SMB accounts.

This shift towards a partner-led motion is designed to drive high-volume new logo acquisition. And you will continue to see this strategy deployed in other geographies of the world in the future.

Finally, our tech alliances with hyperscalers and Frontier Labs. They act as a force multiplier across all segments. They ensure that Zscaler remains at the forefront of AI innovation regardless of company size.

And in FY '26, we captured a record $1.5 billion in TCV through marketplaces and saw over 125% year-on-year growth in TCV through our tech alliances. And our final growth lever, this is so important for us. This is a huge focus for our sales teams and for the company as a whole. And that is customer obsession. So many of you may recognize this Infinity loop from our 2024 transition towards an account-centric selling model.

It represents a continuous cycle where presales alignment and post-sales success are inseparable. On the buy side, we focus on deepening CXO engagement and differentiating the Zscaler brand to ensure we are not just a line item, but a strategic partner.

Our teams are highly engaged, and this is a consistent point of feedback we hear from customers when they choose to partner with us. And on the own side, we introduced Z-Flex 6 quarters ago, and the results have been incredible. Ultimately, Z-Flex allows our customers to do larger and longer-term deals with us to test newer offerings and to swap products at their own pace.

This has led to more strategic engagements with customers and prospects. So it really is a win-win all around. We've seen tremendous momentum with Z-Flex bookings increasing over 60% quarter-over-quarter in Q4. And Kevin will go into these results in more detail when he speaks.

In addition, by leveraging partners for expert deployment, we ensure that first year adoption is seamless and scales as our customer base expands. And the big takeaway here is this. We have evolved the way we go to market to ensure customer success is at the center of everything we do.

Let me give you an example of how focusing on customer success pays off. I want to highlight a public sector customer where early adoption led to rapid scaling, thanks to our focus on customer success. We began the journey with this customer in FY '23 with a small land with our foundational Zero Trust SASE for users and data security solutions. Growing the ARR to $1 million in FY '24 as we expanded in both of those categories.

Importantly, because we established Zscaler as the architectural standard early on, we were able to rapidly repeat this success across more than 100 agencies in this country. Most importantly, we layered on agentic SecOps, providing these agencies with a single pane of glass for risk-based prioritization within their specific business context.

This drove our footprint to 16 million, a 170x increase from where we landed initially, and we have even more upsell opportunities in our sights. So on the next slide, our go-to-market growth levers. They are all underpinned by a relentless focus on go-to-market productivity.

One of the most encouraging indicators of our go-to-market health is the consistent double-digit growth in sales productivity we have been able to deliver over the last 2 years.

I have been deeply involved in the team that is shaping and driving this transformation. And as you can see, the productivity is trending. And we aren't just growing our headcount, we are making our AEs more effective.

This is driven by 4 specific strategic levers. One, I mentioned customer-centric selling, the scale of our platform. I talked about partners being force multipliers and the intentional work we've done with territory design and optimization.

And as we enter FY '27, continuing this upward trajectory is a top priority of mine. By ensuring our territories are healthy and our partners are activated, we are building highly efficient sales machines that can sustain significant growth while maintaining sales productivity. And I'm going to very briefly touch on how we're using AI internally in our go-to-market strategy.

We think about how we use AI initially is being -- it should be fundamentally reimagined. We aren't just talking about productivity. We are embedding AI across the entire life cycle to drive velocity and quality. In pipeline, we're moving from manual prospecting to agentic outreach, targeting the right white space recommendations at the right moment when the customer is ready to engage, and the results are notable. We've seen a 2x higher reply rate to these e-mails, demonstrating our ability to reach the customers at the right time.

And in sales productivity, we launched an AI copilot that has become a force multiplier, handling everything from account intelligence and building complex business value cases, and we are now extending it to agentic quoting. This allows our reps to spend less time on administration and more time on high-value strategic selling.

And in customer satisfaction, our AI customer support agent has driven a 40% reduction in ticket volume, freeing up our human resources to focus on the most complex customer challenges.

Taken together, these are the primary reasons our go-to-market effectiveness continues to outpace the market. And finally, key takeaways that I hope you heard. I cannot be more excited about stepping into the CRO role at this moment at this time with this trajectory.

Zscaler was built to capitalize on this once-in-a-lifetime AI era. Our Zero Trust architecture, it is highly, highly differentiated. And our platform continues to expand in ways that are clearly resonating in the market. We have taken action to enhance our ability to penetrate the new logo opportunity we have, and we're leveraging our partner market as a force multiplier.

And we are doubling down on all the reasons customers are obsessed with Zscaler in the first place. Our team, we are laser-focused on executing, and we are laser-focused on achieving our financial objectives. And Kevin, Kevin is going to walk you through these financial objectives. Thank you.

All right. Hello, everybody. Thank you for being here, and thank you for sticking with us. It's been a lot. Okay. I'd like to start by briefly recapping what you've heard today. We've talked about the structural steps we've taken to increase our TAM, strengthen our product portfolio as we enter the AI era and enhance our go-to-market engine.

These are each important pieces of the growth opportunity ahead of us, and I'm going to walk you through how all of this fits together from a financial point of view. I'm going to focus on 3 areas. First, our multiple growth engines that we believe position us to deliver continued compounding ARR growth.

Second, the success we are having in driving broader adoption of our platform, which is supporting new logo growth and upsell opportunities. I'll provide some data that helps make that clear. And finally, our updated long-term financial framework. Let's start with a quick look back at our track record. We're entering our next phase with a proven track record of growth at scale.

In fiscal '26, revenue grew by 25% year-over-year, while ARR also grew by 25% -- excluding Red Canary, revenue and ARR both grew 20%, and we delivered 14% net new ARR growth for the year.

That momentum gives us a strong foundation to capitalize on the significant growth opportunities ahead. As we've scaled the top line, we've also continued to improve the profitability of the business.

In fiscal '26, non-GAAP operating margin reached a record approximately 23%. Free cash flow margin was also 23%, and we reached a Rule of 49 performance. Since fiscal '23, we've expanded operating margin by approximately 8 points.

So alongside strong growth, we are driving operating leverage and building a business with a meaningful profitability profile. Okay. As you heard from Jay earlier this morning, we are operating in a large $220 billion addressable market, which you can see here map to our platform. We used this Investor Day as an opportunity to refresh our market forecast, and we are now leveraging industry forecasts from third-party market research firms as the inputs to our TAM.

And in addition, you heard from Adam and Dhawal today, we continue to innovate and introduce new products, which has significantly expanded the markets we play in today. We strongly believe our platform is perfectly positioned to capture a growing share of this market. Others have presented a similar view of our platform this morning, but I want to drill down on how we organize the platform from a financial point of view.

Fundamentally, our platform consists of 4 product categories with different levels of maturity and growth profiles. Zero Trust SASE Everywhere is the foundation. It accounts for approximately $3 billion of total ARR and has grown at a 20% CAGR over the past 3 years.

Since this is the first time we're breaking out the platform in this way, I also wanted to double-click on 2 distinct offerings within this category. Zero Trust SASE for users, which includes ZIA, ZPA, ZDX and Zero Trust Browser, is our largest business.

Zero Trust Branch and Cloud are much newer offerings on a relative basis and growing considerably faster at 75% ARR CAGR over the past 3 years. Combined, these have now reached $250 million of ARR, an encouraging milestone and one that reinforces our confidence in the long-term growth potential of the solution.

The second product category is data security, which has grown at a more than 30% CAGR over the last 3 years. Under this new view of the platform, we have shifted Zero Trust Browser into Zero Trust SASE for users to better reflect how we go to market with this offering and as a result, have updated our data security ARR accordingly.

The 2 newer product categories in our portfolio are security for AI and Agentic SecOps. Within security for AI, we ended the year with approximately $70 million in ARR, and incredibly strong momentum as evidenced by the growth we've been seeing. We are very pleased with our results to date, adding to our optimism about the future. Finally, looking at Agentic SecOps, we just launched our new integrated offering. So the historical performance here is less important than where we're headed.

Across this portfolio, we see 3 growth engines that will continue to propel our growth going forward: Zero Trust SASE for Branch and Cloud, data security and security for AI. Combined, these 3 growth engines delivered 60% ARR growth year-over-year in fiscal '26.

You can expect that we will continue to update you on the performance of these growth engines as a group quarterly and on an individual basis annually going forward. I want to spend a few minutes double-clicking on each part of our platform, starting first with our Zero Trust SASE Everywhere portfolio.

We continue to see healthy growth from our Zero Trust SASE for user solution despite running into the law of large numbers. While today, these solutions form the foundation of our portfolio, our business continues to shift more towards non-user products, and our goal is ultimately for our growth to not be tied to seats.

We are making great strides in this regard. As we shared at earnings, 30% of our new and upsell ACV in fiscal '26 was from non-seat-based or metered products. Branch and Cloud are key contributors to this, and both have been strong contributors to our growth with Zero Trust Branch and Cloud ARR growth exceeding 60% for each and every quarter over the last 2 years.

Even more encouragingly, we still see significant white space opportunities for these solutions with just 10% penetration in the Global 2000. Looking holistically at our Zero Trust SASE solution, our customers who are Zero Trust SASE Everywhere enterprises or those who have purchased ZIA, ZPA, branch and cloud are also our most valuable customers. These customers delivered over 5x the ARR of customers who did not have branch or cloud, and they have 121% NRR or 6 points higher than our overall NRR.

Next, our data security portfolio has become an increasingly important contributor as enterprises confront a new wave of data exfiltration risks. Data security ARR accelerated to 34% in fiscal '26 with bookings increasing more than 40% year-over-year.

As you heard from Dhawal, our data security solution is the most comprehensive in the market with products that span both data in motion and data at rest. We have seen over 40% ARR growth from customers who deploy multiple data security products. And today, less than 20% of Zscaler customers have multiple products, representing a meaningful upsell opportunity as we look ahead.

Finally, our new security for AI offering is becoming an increasingly meaningful contributor to our performance. In the fourth quarter of fiscal '26, security for AI bookings increased more than 50% sequentially on top of a strong Q3, and our pipeline increased more than 75% sequentially.

This is exciting progress, and we are still in early days with over 65% of our security for AI ACV in Q4 coming from products that have only been in market less than 2 quarters. This gives us significant confidence that these products will become increasingly meaningful contributors as adoption grows. And as more products within the portfolio enter general availability later in fiscal '27, including our Zero Trust Exchange for agents, we expect even further momentum.

As we've expanded our product portfolio, we've naturally created more ways for customers to land with Zscaler beyond foundational Zero Trust SASE for user offerings. We are seeing tremendous interest in our newer products, and they are becoming a meaningful part of our new logo ARR.

In fact, roughly 1/3 of our new logo ARR today is being driven by these offerings. This is up 3x versus 2023, demonstrating that Zscaler's platform value proposition is increasingly resonating with customers. Once customers are on our platform, we see consistently strong retention and healthy expansion.

Net revenue retention has remained consistent around 115%, while gross revenue retention remains in the mid-90s. This reflects both the mission-critical nature of our solutions and expansion opportunities we are unlocking with our customers. The power of this expansion story is evident in our largest customers. As customers expand their investment with Zscaler, we see growth accelerate. In fact, our $10 million ARR plus customers have grown ARR 9 points faster than our $1 million to $10 million ARR customers over the last 3 years.

And among those, our top 10 customers are growing even faster still. That is powerful validation of the platform strategy and the results that we are providing to our customers. As you heard from Ross earlier today, one of the tools helping us accelerate platform adoption and drive customer expansion is Z-Flex.

By giving customers the flexibility to consume capabilities across the platform, Z-Flex lowers friction in the buying process and encourages broader adoption from day 1. And these results have been compelling with 30% ARR uplift in fiscal '26 from customers with Z-Flex.

And as I mentioned on our Q4 earnings call, Z-Flex bookings exceeded $1.7 billion in fiscal '26, which equates to nearly $500 million in ARR. This is how we will be sharing Z-Flex progress going forward. We've included both Z-Flex TCV and ARR in this slide, so you have the history.

Importantly, even with the excess of expanding within our customer base, we still see significant opportunity ahead. Today, our approximate $3.8 billion of ARR represents only a fraction of the $19 billion plus of addressable spend within our existing customers.

That leaves over $15 billion of white space opportunity sitting inside our installed base. This includes our Zero Trust for SASE user solution, where we see approximately $6 billion of expansion opportunity. As you heard from Adam, this includes the meaningful expansion opportunity in our installed base from ZPA with 30% of ZIA users today who have not yet adopted ZPA.

But more broadly, we also see significant expansion opportunities with our newer solutions. The fact that over 40% of our installed base has not yet purchased any of our solutions beyond Zero Trust SASE for user underscores the magnitude of this runway.

Bringing all this together, I wanted to walk through a customer journey that reflects what we are increasingly seeing across our customer base. In this example, our relationship with this Fortune 500 technology company began with a relatively modest deployment focused on Zero Trust SASE for users and data security. Over time, this customer expanded into additional use cases, including branch, cloud and SecOps.

We ultimately saw this customer commit to a Z-Flex deal in fiscal '25, enabling them to adopt the full platform and accelerate deployment across the organization. Then in fiscal '26, following the launch of our security for AI portfolio, they expanded their deployment with us to include these new solutions and as a result, further expanded their investment with Zscaler.

Over 6 years, this customer grew their ARR roughly 30x with ARR increasing at a CAGR of approximately 74% to $20 million-plus ARR by the end of fiscal '26. While, of course, every customer journey is different, this demonstrates the growth drivers we see as customers move from an initial deployment to a broader platform adoption.

As we look ahead, it is the power of our customer growth flywheel and the meaningful runway for new logo growth that Ross shared that reinforces our confidence in the opportunities ahead. We are reaffirming our fiscal '27 guidance, which you can see here on the slide.

Beyond fiscal '27, where do we see the business going? You've heard us speak about our aspiration to reach $10 billion in ARR. As we execute on this trajectory, we see a clear path to grow ARR organically to over $8 billion in ARR by fiscal '31, more than doubling our current ARR within the next 5 years.

How will we get there? We expect continued strong growth from these -- from our 3 growth engines: security for AI, data security and Zero Trust branch and cloud. At the same time, our Zero Trust architecture is more important than ever, and we expect our foundational Zero Trust SASE for users offering to remain strong, delivering sustained double-digit growth.

And as you heard today, we see significant opportunity driven by the adoption of Agentic AI. This is still a new and emerging area, and so we're not asking you to underwrite this upside today, but we believe it has the potential to be a meaningful accelerant across every part of our business.

Let me talk you through this scenario. While this is not in guidance, it is intended to show the potential of the business, assuming a more rapid adoption of Agentic AI. Our Zero Trust architecture uniquely protects against the threats created by Agentic AI, which we believe will accelerate demand for our Zero Trust SASE Everywhere solution across users, branches and workloads.

With the increasing prevalence of AI, data security and security for AI become even more critical as enterprises look to safely deploy Agentic AI at scale. And with AI-driven threats operating at machine speed, we believe our Agentic SecOps offering is poised to disrupt the SecOps market, opening a meaningful adjacent growth opportunity.

We believe these tailwinds could be significant over the next 2 years and could contribute nearly $2 billion of upside to our ARR. This would put us on a path to reaching $10 billion in ARR by fiscal '31 with each part of our business growing at a faster rate than in our base case. Let's now turn to our financial framework for -- this assumes over $8 billion in ARR by fiscal '31. I'll start with gross margins.

As you can see, our gross margins have consistently been around 80%, and we expect this to continue. We have been driving productivity in our go-to-market organization, and we expect this to continue as well with sales and marketing as a percentage of revenue trending down. For R&D, we expect modest leverage from our more mature products while continuing to invest in new innovations. And in G&A, we see this ticking down modestly from AI adoption and automation. All of this leads to continued operating margin expansion by fiscal '31.

Regarding free cash flow, we expect margins to be above operating margin. However, there are some timing considerations given the current memory price backdrop. Specifically, we are expecting CapEx as a percentage of revenue to remain elevated until fiscal '28, then normalize back towards high single digits as manufacturing capacity comes online and memory and component prices decline, which aligns with the forecast from Gartner and IDC.

Finally, I want to touch a little bit on stock-based compensation as this is a key factor in driving GAAP profitability. Over the long term, we are focused on bringing this in line with our peer group average.

This excludes any future M&A, which has historically contributed 1 to 2 points. With this in mind, we are always evaluating the best use of capital to maximize shareholder value, and our capital allocation framework is clear.

Our first priority is reinvesting in the business to drive continued innovation, resiliency and product expansion to drive sustained long-term growth. Second, we deploy capital into M&A, focusing primarily on technology and talent tuck-ins to strengthen the platform.

And third, we are committed to maintaining a strong balance sheet to preserve financial and operational flexibility. To wrap up, I'd like to leave you with a few points. First, we are operating in a very large and expanding market, and our innovation continues to broaden the opportunity in front of us.

Second, we have a demonstrated ability to translate that opportunity into profitable growth at scale. And third, with multiple growth vectors, significant white space in our target market and the tailwinds from Agentic AI across the platform, we are positioned to drive durable and compounding ARR growth for years to come with both new and existing customers.

Taken together, we have a strong foundation, clear visibility into the path ahead and confidence in our ability to create long-term shareholder value. Thank you.

Okay. All right. Up next, we have our Q&A session. So I'd like to bring all of our speakers up to the stage for this section.

Okay. I see the hands up. So when we call on you, just remember to state your name and your company name for the webcast. Really, really appreciate that. Brad, I think you had your hand up first.

It's one of my skills. Brad Zelnick, Deutsche Bank. Thank you for an amazing day, really great and compelling presentation.

I want to direct my question at Ross and maybe for Jay to chime in as well. It seems there's been a lot of change in the sales organization. Great to see you stepping into the role as well. I guess as we think about the underlying drivers of turnover that you've seen, what it is that there is -- there might be to know why we shouldn't be concerned by any ongoing disruption as a result of these changes? And then I'd also love to hear from you what's new, like from going from one leader to the next, usually, there's your fingerprints that you're going to want to put on things.

As I look at the presentation you gave today, it sounds like with the 75% of the market that you haven't reached yet, it would seem like there's a real opportunity to go further down market. And related to that, I'm just wondering the resources put behind that versus maybe not then allocating enough towards this amazing $2 billion AI opportunity that we see and how you balance that. So I know a few different questions, but all related.

4-part question, I think. Let me see if I can remember all those. So I'll start with the latter part. So new logos, as you heard, are a real emphasis for us. And with the expanding platform and things that we've done in our go-to-market machine around focus and incentivizing our teams, I feel like we'll accelerate there.

As we look at the down market opportunity, that is not a discrete focus for us. It is an opportunity, and that's why we're leveraging relationships like Carahsoft that I mentioned. We need to drive velocity and scale, and that's what we're focused on there.

We will remain very focused on enterprise and majors. We do have an opportunity in that space. We're not ignoring it, but that's where we're really going to leverage our partners. You asked what I would do differently. Yes, you led the witness a little bit in that new logos are very important for us. We're driving verticals.

I think I talked about vertical strategy. We have a lot of opportunity in public sector, international, as well as sovereign cloud. So we are really leaning in there. And then obviously, the platform opportunity. I have a long history of selling platforms. I know how to do it. I know how to lead those teams.

And so we're very focused on that. And then I think coming around to your first question, and Jay, I'd encourage you to chime in as well. I mean, yes, there have been some changes, and they've all kind of been unrelated to each other, frankly. So there's not any individual macro thing. We've got a strong leadership bench.

We have our leadership team in place. I'm very happy. And as I've mentioned multiple times, we are really focused just on execution.

If I may add, you saw us reiterating the guidance. That tells us the confidence we have. You also saw us give long-term guidance as well that could tell you that we are very confident on the opportunity.

The overall platform need for cyber, our role as being the linchpin is fundamental. The sales process and strategy we put in place is doing well.

We just need to keep on accelerating on it. And I think we have a strong leadership team, and there will be some changes from time to time, very confident.

Okay. We're going to go to Fatima, and then we'll head over to Joe. So Fatima and Joe in the first row. Can we get a mic over to Fatima in the middle? Can you keep your hand up for a sec? Webcast, thank you.

Fatima Boolani from Citi. Thank you so much for such an insightful set of comments and presentations. Dhawal or Adam, for you. Dhawal, you mentioned something that was really interesting to me in your session just around this intersectionality of security for AI and data security.

Now those 2 seem to me very interrelated because you can't have secure AI usage without having a handle on your data assets. And the data attack surface or the data risk surface, as you pointed out, is absolutely metastasizing.

So why is there still a distinction from a product standpoint if presumably data security is going to be a conduit for security for AI and vice versa? And then as a related question to Ross, why not unleash Z-Flex to the entire base? What are some of the limiting reagents? Because if you are going full force from a platform perspective, you talked about only 9 of the 30 individualized SKUs are being taken. Why not just tear up kind of the traditional motion? You've seen the success with Zscaler -- I'm sorry, Z-Flex and go all hands on deck.

So probably I'll start with the first part. When we look at the AI security broadly, right, or security for AI, data security is a very strong tailwind as I was explaining the numbers.

But security for AI is broader than that. A lot of new cyber-attacks that we are seeing. Think about how users are being tricked to -- get phished these days. It's happening through prompts, right? Prompt injection, jailbreaking is happening through that as well.

So there is definitely more to it. But -- and as you saw, securing AI also needs to secure AI applications and infrastructure that you're using from what goes into your models. And data security then becomes a key pillar on that. You're absolutely right that AI agents are not just consuming data.

Our models are not just consuming data, they transform the data. They also leave remnants of this data in multiple places. And this is why we've been classifying data, and we have a strong DLP business that has grown significantly. But the challenges that come with data security in the AI world are also different.

And this was one of our core thesis for acquiring Symmetry Systems that we can discover data everywhere. We can discover AI everywhere. What is really hard in this world is understanding the relationship between data and how that data goes into AI and how identities are [ playing pay ] with very excessive permissions and role.

And that problem is getting compounded with stat that Jay showed, there are 75 agents for every single user. So the short answer, in my opinion, is like yes, data security is a big part of it, but AI security in itself is much broader. That's why we are keeping it that way, but we are bringing it together. Adam?

Yes. Maybe just one quick add on that. The -- this is why data security is part of the platform, and it's a consistent capability across everything. I didn't spend much time talking about it, but it's foundational in everything we're doing in Zero Trust SASE as well. What we've been recognizing is where does that need to evolve.

And primarily, AI has just served as a major accelerator of if we knew this is important before a customer knew, now it's really, really important. And then you're seeing the specific pieces around data security is always about find the data, classify it and then enforce policy. What we've learned in AI, there are new elements of what that policy looks like. And it's broadly applicable, not just for AI, but AI is driving that, and that's why we're doing it at a platform level, so we can use it across every one of our product areas.

Ross?

Z-Flex, and Kevin, chime in if you have any thoughts. So that is a growth enabler for us. We don't have any limiters on -- I mean, there's some obvious baseline you need to spend x minimum. But yes, Z-Flex is kind of a core motion for us.

And we did give an example where a customer actually in the case study is able to take the entire AI security portfolio through Z-Flex as well.

With so many questions coming up, we're going to keep our answers short. So we can take more questions.

Thank you, Jay. He's taking my role. Now we're with Joe.

Okay. So I have a question for -- Joe Gallo of Jefferies. Thanks for this. A question for Adam, primarily on SecOps. So I know it's early, but who are you competing with? Like I assume in your existing customers, you are the network security platform.

So I assume you're competing with the endpoint-centric platform. So I'm just curious, your right to win there because those vendors also have a lot of data.

They certainly do. And I think the data gravity is the main point. Every customer that we work with when they're figuring out SecOps, they're going to orient around some center of gravity.

And in some cases, in many cases, that's going to be us. It can also be other players that they have. So the solutions -- anyone who has a SecOps solution, if they're not they don't have that data gravity. They're a new agentic SecOps start-up. I don't think they have that easier right to win compared to a Zscaler or a larger platform play.

SecOps has become a more and more of a platform play. That's why I think we have that right and that shot to be competing in that space with the telemetry of the data we have and that agent-first approach that we're taking. I think Kim likes to bring it up.

I have this comment or say, we get sometimes a second mover advantage. We weren't first to this space. We recognize that. But the benefit is we didn't spend all of our time building a solution around log collection and storage. We built our time -- we built it all around an agent-first approach.

Adam, we are first mover when it comes to solution built for agents. Others build solutions for humans and try to add agent to it. That's point one, the second short answer would be data. We got full in-line data. We got endpoint data. We've got cloud workload data. We also got branch devices data. We have better telemetry than anybody else there.

And we bring in third parties. We'll go Ittai, and we'll do you too. Adam? Can we get a mic right here in the front? Thank you.

Thank you very much. Ittai Kidron from Oppenheimer. Great presentation. Really appreciate the disclosure.

A couple of questions for me for you, Adam, first, on the technology side. You talked about the importance of context. I guess why not own the identity side of the equation? It sounds like you are drawing from identity in order to create context, some of your -- your biggest competitor has decided to own that layer.

We already see some data security companies moving into the identity side to create that context more likely, more closely linked. I guess the question is, why not own that category instead of partner over there and potentially end up with no partners if they get acquired.

So that's question one. And for you, Kevin, on the financial side of the equation, I appreciate the long-term targets on fiscal '31. I guess if I think about just that $8 billion target, just kind of doing back of the envelope, it sounds like net new ARR needs to be growing around the 10% plus/minus on a consistent basis for you to reach that target. That's already well ahead of your target right now for fiscal '27.

So are you already raising the numbers? Or should we look for a little bit -- are you going to grow into this fiscal '31 number? Or how should I -- how do we think about that?

Yes. I guess I'll go first. So the 17% CAGR is in line with the midpoint of the guidance for this year. So that would be my direction there. As we think about the opportunity to $8 billion and potentially $10 billion, let's remember, we have a huge opportunity in a giant market. We've talked about new logos. We've talked about continued extension of the platform. We've talked about moving even beyond into adjacent SecOps.

So I think there's a lot of opportunity and a lot of reason to be really excited about the opportunity. But that's how I would think about the longer-term model.

If I may add, I'm only focused on $10 billion number.

Adam, do you want to finish off the question?

Let me try to. So on the -- from the identity piece, needing -- leveraging identity as context is part and parcel to everything we've done in Zero Trust that it requires that.

It does not require you necessarily owning it. Customers use identity for lots of different things, and they have identity players and components that bridge well beyond security. So if you want to get into that space, you have to sign up to do all of those pieces. So we've been very successful leveraging it as key context. The other piece is identity players are not typically in line. They're in line for one piece of it. They're not an in-line processing engine, which is what we are. That's where we are focused. We're focused on leveraging that context. It's a relationship for sure. It's getting a lot of attention, but I don't see the burning need to say we have to own it because I think that brings a lot of other distraction and you saw what we have to go after.

We have a lot in front of us, and I think focus is going to be very important as well.

If I may add, in the agentic world, identity, basic identity, who are you is coming from the party that's creating agents. That's the starting point. Identity doesn't give any more context than who you are, who are you.

The rest of the context come from sitting in line. We have lots of context today for user identity, as Dhawal talked about without having the core identity. Would I go and spend $25 billion and hoping that this business will become agentic identity? Not really. Will I work with Microsoft and Azure and AWS and Google of the world and take identity from them, add context to it, add intent to it. That's a winning formula.

Adam?

Awesome. Adam Borg with Stifel. Thanks for the time and really appreciate the conversation today. Maybe for us, so talking about the new logo opportunity, especially down market, talk about the confidence you have there, especially given investor perception, it's a lot more competitive down market.

So like what changes would need to be made, if any, on the pricing and packaging side to be more competitive there?

Sure. So at a high level and keep this brief, a couple of things. A, we are looking at more simple bundles like T-shirt sizes that we are then leveraging our partner ecosystem.

I talked about Carahsoft, and we are looking at doing similar relationships in other geographies. And so having those simplified bundles through a partner machine that can leverage and touch all those organizations that we just don't have the human capacity to do is why I feel confident that we can grow the new logos in our commercial and SMB space.

I mean, an interesting challenge we have is the upsell opportunity is so big. The new logo opportunity is so big. We are trying to do a balancing act rather than saying A versus B.

And the only thing I would add to it is, remember, we have about 4,600 of what we've determined to be about 20,000 of the largest customers.

So there is a 75% white space, if you will, relative to new logos. So there is a large opportunity for us to continue to replicate the success that we've had with existing customers into the 75% that we've yet to serve.

Okay. Let's go back to John in the corner, and then Saket up in the front.

Good eyes, Kim, way back here. It's John DiFucci from Guggenheim. I guess, I want to go back to Fatima's question because AI does look like a discrete opportunity at this point out there.

And that's how all cyber companies are addressing it. But if AI becomes part of everything, it is much broader, as Dhawal mentioned earlier, doesn't it just become part of everything you do and everything everybody does? And if that's the case, that's -- that $2 billion number you put out there, to me, I just wonder how we should be thinking about that and frankly, help the risk in it?

And then just a quick follow-up for Kevin. Why in '26 did the delta between free cash flow and operating cash flow narrow so much in '26. And I don't think it was just that CapEx.

Do you want to start?

Yes, I'm happy to go with the last one. John, we did see a pretty significant tick up in CapEx as a result of pricing and supply chain-related items. The other deviations is really just timing.

So maybe I can start on the AI and data side of it. We're going to keep the answer short this time. Yes, AI and data are interrelated. You saw the presentation for the first time I've seen any presentation with AI and data is sitting together. Discovery of AI, data, risk of AI and data, who accesses what, they're all related in our solution.

So we are selling it as a combined solution, but also giving options to customers to buy individual solutions. There are separate buying centers for data security today. AI is separate, but they're coming together. As time goes on, you're going to see more and more stuff being sold together.

I'm sorry, it's more than just data. If you think of endpoint, you think of identity, you think of -- I mean, they're all related to data, but it's more than just data. It's process. It's things happening, everything happening out there, right? I'm just concerned about the $2 billion out there because I have -- I think you guys are doing a great job, but it seems like some risk to put that out there.

Lots is happening, but we need to figure out. Our job is to make sure right entity get access right entity, so things don't get compromised and data doesn't leak out.

So we're not trying to figure and fix every workflow in the company. Right security, the right access for that identity plays a role, inspection plays a role. And AI is the main source and data is one of the key pieces to look at. So we can take it more offline.

Thank you, John. Saket, right up here in front.

Saket Kalia at Barclays. Great, great session as always. So Jay, this question is for you, right? The message on the mix shift, right, as part of this long-term plan is very clear, right? There's about 25% of ARR, give or take, right? That will become a bigger mix by 2031, right?

But what about the user part, right, which is right now is the vast majority. I think let's just assume the $8 billion right now, right, just for a base case, that will still be the majority, I think, right, by 2031, you correct me if I'm wrong.

I just want to ask the question, can the user part of the business grow faster than double digits? And a question that hasn't been asked yet today, that I think is on all of our minds is, what about competition? Can we just do a level set on competition in the user market?

So if you think about competition, first of all, we pioneered securing users of Zero Trust. Largely, competition is now trying to secure users without Zero Trust.

When you guys talk about all the firewall vendors as a competition, none of them are Zero Trust even for users. All this SASE and SSE is still firewalls and VPNs running in the cloud. That's the starting point.

And they cannot afford to change to Zero Trust architecture because of cannibalization that's going to cause because all the firewalls go away. Our customers are telling us that they need to take care of that. So number one, the next part, we are not sitting end users.

Customers are asking for Zero Trust everywhere. I'm not going and saying, buy my user stack. I'm going saying you need Zero Trust everywhere. Users, branches, devices, cloud workloads and next becomes agents.

You need to look at the solution and customers buying us, holistic story, user is one part of it. It's an important part of it. But I think if you ask me what piece will be the biggest piece, it's too early to tell, but I'll tell you, agents will become the biggest piece because that number will grow significantly. That's even more risk than users.

And we are so well positioned to sit in line, understand how to deal with agentic traffic. And also, we've taken one other approach. A lot of stuff will happen in the cloud and exchange. Some will happen on the endpoint, and they've all talked about having a solution that takes your endpoint and exchange both. We are pretty unique in that area.

Okay. We'll do Peter in the second row and then up to the front for Roger.

Peter Levine, Evercore. A follow-up question to Saket's. When you think about competition, your predecessor, what are you keeping? What are you changing? So if you think about you now stepping in the role, like what are you thinking about? Just help us level set some of the changes, anything that you're thinking about that worked? And then second, Kevin, to John's question, I was going to ask it similar, but maybe if you think about the $2 billion, is that a base case? Is that like the bull case, bear case? Like how did you come up with that number and the realistic based on the products you have today, like, what does that fit?

Yes. So I'm happy to kick it off. Look, we outlined a base case that gets us to $8 billion by fiscal '31, more than double the ARR.

And I kind of articulated through how we see the growth opportunities to achieve that case. The $10 billion is really a scenario where AI happens faster and in a much more meaningful way, and it happens across the totality of what we offer, right? One of the things that you've heard from us this entire morning is we believe that the architectural foundation of what we offer is uniquely going to be driven by AI, right? When you think about all of the different threats and exposures that exist, the ability to reduce your tax surface and eliminate lateral movement is just fundamental to how you secure AI.

And so we think that everything in the platform will ultimately benefit. So I'm not trying to put a $2 billion bogey on just AI. It really is the momentum that likely could play out with AI and the benefits across the totality of the platform as a result of that. I don't know if Jay.

If I may add, I think the demand is growing, everything is growing so rapidly. And we will not be constrained by market size. I don't believe we'll be constrained by competition. The main thing will be how well we execute as a company.

Our execution on internal products and go-to-market is really what we are looking at. And that's why we've given you the targets we're giving you.

Yes. And on your first question, a great segue, Jay. What I'm thinking about is execution. I've been at Zscaler 3 years. I've been part of every strategic conversation. I was part of the annual planning.

So we have a plan and it's time to go execute it. And the elements of that plan, we've talked about, I think, a bunch, new logos, platform expansion, international public sector, sovereign cloud, maximizing our specialty sales team.

Those are the things I'm focused on, and it's really about execution versus a change in strategy because I think we have the right strategy, and it's time to put the foot on the gas.

Just to add on, the question was, what changes. As Ross has said, we aren't expecting any meaningful changes. The approach is working. We just need to keep on executing with focus with a strong team, strong bench.

Changes from time to time are a good thing. So I'm very confident in the leadership and go-to-market, and we will execute and we'll deliver on our numbers.

Roger, right here at front.

Cool, thank you. Roger Boyd with UBS. Jay, it felt like one of the themes today was the strength that new products are having in driving new logos. And you talked about 40% of Zero Trust branch customers that are new to Zscaler, the 350,000 branch win that you had and the ability to land with AI Gateway.

It seems like a lot of this may be product-led. You talked about the differentiation you have compared to SD-WAN as well as AI Gateway.

But can you talk about like how these customers are finding Zscaler? And maybe for Ross, how do you think about building like sustainable sales motions to go after these non-ZIA and ZPA lands? And how does that kind of factor in kind of the next 5 years?

So your first part, Zero Trust branch, it's growing very rapidly. And the new logo -- actually, new logos aren't coming because of just the branch.

New logos are coming because when the customer buys Zero Trust for users, they want Zero Trust branch at the same time. So the need is getting bigger, and this is helping us win bigger deals and a bigger part of the solution. You've always seen us do early on.

Remember, we used to have ZIA just a business enterprise. We combined the bundle, ZIA, ZPA. It's natural when customers want a bigger part of the platform, Zero Trust users and branches happening together. We'll see as we're starting to see more and more cloud verticals happening together.

Our goal is to sell the platform since it's integrated platform, it's working.

And I think your other question is around sales plays?

Just when you think about kind of investing around kind of landing or something like AI Gateway?

Very important. And so Adam, maybe you can chime in. But -- so we work with the product teams very closely to develop these sales plays and make these sales plays consumable for our field as well as our customers, pardon me.

And then we work with our marketing team to launch that out there. So we've got roughly 5 to 6 sales plays that our teams have been enabled on, which are at a high-level conversation and then they bring in the real technical people to get under the covers.

Yes, very quick add on that. And the reason why I highlighted this in my presentation. So yes, I'm glad you did pick up on it. I'm 2 years at Zscaler. This is -- that was not as high a priority of a focus.

So that did involve, you have to change your -- you have to adjust your products so that they are easier to land and more focused, and that's what we did.

And so that creates this new opportunity for messaging and how to explain to customers where you can start with us and then the selling motion of that. And so that -- this is really the year where we're driving that with force. So we're excited to see where that plays out. That's why we wanted to articulate that to this group.

Okay. We'll go to the second row here, Shrenik and then Gray, and then we'll move down to Gregg.

Shrenik Kothari from Baird. So all of you guys did reinforce the reacceleration potential, the reaction embedded in the long-term framework. Well, it did not go too deep into what is still relatively the undermonetized or opportunity to monetize the Zero Trust Exchange for Agents, right, which is going to be indexed towards traffic and agent transactions. So just how should we think about that agentic monetization evolving? And then how is that contemplated in the long-term framework beyond just kind of reacceleration through new logos and users and seats?

I'll start. So if there's #1 thing that's holding enterprises back from fully rolling out agents is the security policy governance concern, okay? And it's not easy to do. Everyone seems to talk, when I got a control plane, I got this in plan.

How do you even get the data, right? And agents can be -- the control plane can be sitting on an EDR endpoint. It can be sitting here. You need to be able to take agentic traffic coming from Salesforce, coming from Snowflake, Databricks, hyperscalers, endpoint, all these places. We've done some of the same kind of stuff before when we did cloud workloads when we did the users part of it. So we are pretty well positioned.

We have a product that we launched in June. It's an early availability right now. And this is going to take some time, even though there's more interest in the design partners, customers for this product than any other product. When you're sitting in line, you need to understand things, right? There are many things that are nuanced in the agentic world. Intent all those contexts, our team is figuring or working with them. But agentic product, agentic exchange will take some longer time.

The other products, visibility, discovery, access graph, all this stuff is good saving for us. I think we're well prepared. So are we really factoring exactly in what we know about agentic -- this change? It's a little bit too early to say. But early indicators are very positive.

Yes. I'll add to what Jay said. One of the advantage we have is that we already have a footprint where agents are getting deployed. On the endpoint, we have a client that can steer traffic to our AI gateway already, right? Then we have seen in public cloud, neo cloud, edge compute platforms, SaaS services where the agents are embedded, we are able to steer traffic from them.

So a lot of that complexity is what we have saw. Now building the functional layer of features and policies and intent and context is what we are building. So very strong design partnership, and we expect this to grow in a meaningful way.

Let's head to Gray in the second row right there.

Great. Gray Powell with U.S. Bank. So actually, I wanted to follow up on Roger's question earlier. And Jay, you hit on this. But we frequently hear that customers, they want to buy secure service edge and SD-WAN together like in the same motion.

So I guess my question is, what gets customers comfortable with your product over a dedicated SD-WAN product or something that just gets bundled in with network security. And then I mean, it sounds like you're seeing good demand for it. So just -- I'm just curious, like what kind of uplift do you see to the typical Zero Trust for user ACV or whatever metric you want to give when customers take Zero Trust Branch?

So first of all, analysts -- industry analysts have asked us, go into SD-WAN business, right? I said we want disruptive technologies that have no lateral movement. So we refused to really go with SD-WAN market, when everyone kind of said, I'm SD-WAN, I'm SD-WAN. So it took us some time. We build Zero Trust Branch, no lateral movement. It's becoming more and more important today.

When we're rolling out the -- getting the product out last year, or early stage, I was thinking that perhaps 2/3 of our customers will embrace it. Others will still say, "I want my SD-WAN." I can tell you every Zscaler customer I talked to, they're all ready to dump SD-WAN and go with Zero Trust branch. I am surprised, but that's what results are.

And this frontier model world where risks are growing of breaches and lateral movement, SD-WAN is a risk. The only time customers buy SD-WAN now, when we haven't been able to spend time and show them the merits of it. When we engage, we tell them, SD-WAN elimination and having Zero Trust Branch and user is a very natural thing. And in next year, I want to be here and say, Zero Trust Cloud workload is a natural thing. And for the same thing and agents. Being able to do Zero Trust everywhere is the key strength of Zscaler, and we build upon it.

Gregg?

It's Gregg Moskowitz from Mizuho. I thought you did a really good job today outlining your Zero Trust Everywhere solution and specifically the benefits of it, hiding all applications, strong micro segmentation, preventing lateral movement, layering on extensive AI security protection, et cetera. But the SASE market does seem to have become more crowded or minimally noisier, I would say, in recent months, if not last year or 2.

When you speak with customers, Jay, Ross, how much confusion exists? And I think we can -- all of us here can appreciate that strong execution can drive a winning outcome. But over, again, in the past year or so, has it become a bit more difficult to articulate your value proposition to customers and to close new and upsell transactions?

I'll start by saying, look, when we get engaged, we almost always win because we're able to clarify it. It is true that every vendor in security wants to be in SASE because if they aren't in SASE, SASE is a big part of it, then they what are in. SASE eventually, if done right, will consume all the firewall business, okay?

So if really firewall vendors were to do 3 of SASE, they'll be worried about SASE because no firewalls will be needed. But our goal is really not to pigeonhole SASE for users. We bring the story together. Our goal is to show differentiation of branch level, show differential in cloud workload level.

Wherever we go, who offers Zero Trust Everywhere or who offers Zero Trust SASE Everywhere? What do you do in the cloud? It's same old virtual firewalls. Nothing is new. Branch is same old SD-WAN. And in the user level, it's firewall in the cloud as a virtual firewall, it's VPN in the cloud as a virtual VPN. That's not real competition. And I think a lot of bundling that has happened, that will come out. And I'm feeling very bullish and confident, we just need to keep on executing on the sales side.

Jay, I don't have a lot to add. I'm very bullish and confident as well. It's about execution. As Jay mentioned, we are fundamentally different, and it's our job to explain the value in that difference.

Let's go to Eric.

All right. Eric Heath with KeyBanc. A question for Dhawal, I guess. Those was really interesting, the comment you had that you were the only vendor with the context and intent to understand agentic and AI action. So can you just elaborate on that point a little bit more? And maybe help us delineate the lines of where the endpoint visibility ends and where you're visibility picks up? Because obviously, there's a lot of rhetoric that a lot of the agentic activity happens on the endpoint. So help us understand what's incremental in terms of unique data that you can provide and be that enforcement point?

Yes. So I'll start by saying that Context layer in a lot of agent security platform is tied to agent identity only. Going back to what I said earlier, the visibility we are building with connecting the dots of data identity and AI and using that as the context with our gateway and endpoint is unique. So that Context layer is the golden layer. Lot of new AI gateways that are coming in the market are very intent focused or use case focused. I think our -- one of the biggest advantage in that area is, yes, the intent policy stack will become very commoditized over a period of time. But being able to build AI gateway at scale at which we run our cloud is something as a muscle that we, as a company, have built.

So bringing context from a Zero Trust perspective and intent from a functional perspective, when you bring it together, we are seeing that validation coming from our customers from a design perspective that we are unique. Your point around endpoint AI security is good because when we think about where policies can be enforced, the amount of attack surface that gets generated in the endpoint with the actions that AI agentic applications can do on the endpoint with data as well as actions they can take, they actually are becoming hard to control before even it hits the network. You can do a lot of damage within the network itself or before it hits the gateway. That's why being able to meaningfully do when things are happening within a prompt and controlling that kind of skill files that people are downloading, for example.

So those kind of controls could be implemented there. But one thing we have already done is our endpoint AI security product is fully integrated with our gateway. When the traffic needs to leave to the Internet or to an agentic application, that is subject to gateway policy if the agentic action on the endpoint is not triggered. So it's best of both our approaches that we are taking in such things, it's either are.

Hands, 1 more time. Josh, right there.

Joshua Tilton from Wolfe Research. There's been a lot of talk today about reaccelerating new logos. If you look at the other side of that, it's been pretty impressive that you've been able to sustain a mid-teens NRR over the last 3 years. When you look at the FY '31 target, maybe just in the context, that $8 billion, do you still expect to be achieving a mid-teens NRR rate?

I didn't call out a particular longer-term target for net expansion, but it's fair to say that I would expect that we continue to have significant success upselling based on everything that you heard today, the broader platform, et cetera.

Okay. Probably time for one, maybe two more. We'll go over to the second row, aisle. And then I think we have one here, and you'll be our last.

Mike Cikos from Needham, and thanks again for hosting the day. Maybe just building off the building blocks here. But Ross, I know you were saying we're looking at execution. And you had spoken about the hunter territories earlier. Can you just give us -- or elaborate on that further, like where are we? Was that something that we just established at sales kickoff and what's required as far as hiring bodies to stand that up, ramp time to start seeing the fruits of those investments?

Yes. And Jay, if you want to chime in as well, you and I've had a lot of discussions about this. So historically, we have had new logo hunters around the globe, and we have expanded that here in FY '27. That is not our only new logo motion though. We have territories that have a combination, they're hybrid territories with new logos and existing customers, where that makes sense. So we are building out those additional territories. Or they have been built out, I should say, and hiring is almost complete for all of those.

I do -- we've talked about new logos. I do want to come back and remind everybody the white space opportunity we have within our existing customers. We've only got 4,600 out of the 20,000. And then if you look at those 4,600, there's like a $20 billion TAM, and we have $3.8 billion of that. So yes, new logos are important, but we have a tremendous opportunity within our existing customer set.

If I may add, yes, as a part of the fiscal '27 plan, the territories were very well done. We spent a lot of time understanding each territory. What's the mix of new logo versus upset current customers? Because ideally, you want to give a balance of both too many apps, but then having some of the very focused hunter territories. Both are in motion. So I don't want to send a message that we are expecting new logo only from hunter only territories. That's 1 more step but a combination that and existing reps who have both new and upsell opportunities. We expect good results this year.

Yes. Spot on, Jay.

Yes.

Okay. We'll do Junaid, and then I'm going to add one more, Rich Poland, and then we're going to wrap.

Thank you. Junaid Siddiqui, Truist Securities. Jay, we've seen some of your competitors expand deeper into observability. And we're seeing some of the swim lanes convert to security. Where do you see Zscaler uniquely positioned in that conversions? And given your in-line architecture and vision around Agentic SecOps, do you see evolving towards an operational control plane beyond just security?

So we generally got very strong convictions about where we want to go, we don't want to go. And generally, it's not driven by somebody buying a vendor A or vendor B or vendor C. We think our customers want meaningful intiered solutions certain areas. You saw the TAM we had today. It's a massive TAM in a very synergistic area. So do we need to get into observability to look at performance across everything? That's a secondary area, not really. The core markets we shared with you, they're so big, so massive. I would rather dominate those markets rather than trying to expand into other markets.

Where it becomes complementary really quick, and additional though, is, as I said, if we're mission-critical infrastructure for our customer with our Zero Trust SASE Everywhere, understanding the end-to-end experience on that is quite important, and that's why we do have our ZDX, but it's specifically about the experience for things on Zscaler's platform across Zero Trust SASE, not general IT observability. So I think that's the right lane for us to be important, and it is a big value for our customers.

Okay. I'm going to try to squeeze one more in. Rich Poland in the front row.

Thank you, Kim. Rich Poland from Wells Fargo. I think just to kind of bring it all together, I think 1 of the questions that we've gotten from investors is just thinking about SASE and the architectural shift that Zscaler proposes is a large undertaking. And so I think part of the reason why you've seen some of the SD-WAN is because it's kind of almost the path of least resistance in some cases. And so while customers are thinking about AI and all these other things that are going on right now in this current environment. Just kind of how do you think about getting customers that would be new logos or embracing Zero Trust Everywhere over that hump?

I'll start. Actually moving to Zscaler Zero Trust is not a large undertaking. You're seeing customers deploying, say, Zero Trust users in a matter of weeks or a few months. All you need to do is start with a lightweight agent on the endpoint. You don't even touch your network. The traffic starts taking whatever path is available. We define some policies, we can be fully up and running. Now it can take time to dismantle all the stuff that's sitting there. But customers typically get moving with rolled out to Zscaler start getting the benefit of security and then have the benefit of removing the stuff. Now if we get some inertia and pushback, generally, that comes from a lower level. People who own the stuff, a little bit job security, a little bit of inertia, I know this thing, too.

That's why we have very strong relationships, CIOs, CSOs, CTOs, AI offices, who drives change. the leadership does. When they want to embrace AI, they really become a catalyst for us. So we are seeing AI as a catalyst, not to say that Zscaler requires a big infra change. The world has moved to the level where the management is buying our stores. That's why we're seeing a lot of expansions. I mean the customer you heard today and the example we shared with you today, these customers starting from $1 million to $5 million to $10 million to $20 million. This is real, it's happening because of the benefits we offer.

Okay. That's it on Q&A. Jay, do you want to wrap this up?

Yes. Thank you again for joining us today. We hope that today's presentation has given an opportunity to see that, one, Zscaler is the cybersecurity platform for the AI era. Two, AI is the largest tailwind what business has ever seen. And three, we have strengthened our go-to-market engine, and we intend to keep on penetrating it, driving it to achieve our large TAM. We have multiple growth engines that will accelerate our growth to move forward. And we look forward to proving you that we can do it in coming quarters.

Okay. Thank you, Jay, and thanks for all of you for joining us today and making the time. For those of you online, this will end our webcast.

I'm Fatima Boolani, I jointly head up our software research team here. Thank you for joining us in the afternoon session of day 2 of Citi's TMT Conference. I'm very excited to host a fireside chat with the CEO and Founder of Zscaler, Jay Chaudhry. Thank you so much for being here.

Thanks, Fatima.

Jay, I want to start the session and the discussion off by you just setting the stage. You just closed out fiscal '26, which was a pretty transformational and transformative year for you. But before that, I have to say Zscaler is 19 years old this month. It might have been your first baby, but congratulations on that. But a pretty transformational year for you in terms of a lot of activity, a pretty large acquisition. So just to have you set the stage for us and the table for us on the year-end review, the key milestones coming off closing out this year.

So first of all, our platform expanded significantly. Zero Trust everywhere is a big differentiator because you don't want just for users, branches, devices, workloads and our agents. Our portfolio for data security is as comprehensive as anything gets. And then security for AI, actually, a year ago, we had one product that was GenAI security. Now we've got 6 very good products fully integrated. That has exceeded my expectation, the pace at which happened. And we expected it.

We set up a security for AI start-up within Zscaler. It has actually delivered solutions very rapidly. And then the launch of Agentic SecOps that happened today. Great platform on the product side. I think we have a better portfolio platform than anyone else there.

On the sales side, we got some great momentum. We started the journey of transformation about 2, 2.5 years ago when Mike Rich came on board. In that time, we changed the sales team's process quite a bit to be enterprise account focused. And that thing is delivering results. To give you tentative data, you may think that overall ARR growth percentages here. But the real number to look at the growth where we have made changes in net new ARR.

In '25 -- I mean, '24, we basically had very little net new ARR growth. '25, we took it to 7%, fiscal '26 first half was 10%, Q4 was 17%. It's tracking very well. So good momentum on the sales side, good on the customer side. Customer success is very good. Our NPS, Net Promoter Score has been sitting in the 80 to 85 range, very, very proud of it. That's reflected in the amount of upsells we are doing. So very pleased and looking forward to further doing better job in '27.

So product portfolio has expanded materially. The go-to-market motion is now getting to a cruise control or a cruising altitude, if you will. So as we look into fiscal '27, maybe we can spend a little bit of time talking about the security for AI opportunity. That seems to be an area where you're very excited.

When you spend time with customers, what are their most salient and palpable pain points that you're solving today? A lot of vendors are clamoring for the CISO attention, clamoring for the budgetary attention. How are you putting a stake in the ground and wedging yourself in the room and saying, these are the focal points for your architecture today. And I don't have to remind anybody that this has been, again, a pretty consequential year of cybersecurity headlines development, right, post Mythos, post the Hugging Face incident. So how have those conversations shaped the dialogue that you are having with CISOs today that's different from the conversations you were having 6 months ago?

So first of all, it's not just the CISOs. Now CIOs worry about it. The Board worries about it. In fact, after Mythos, most large enterprises had a cyber task force whose job was to do some of these things and then report to the Board on a frequent basis. So here is what's setting us apart in some of these areas.

Think of what the customers are worried about from cyber frontier models. One, even if you don't plan to embrace AI, frontier models still cause a risk because anything that's exposed to the Internet could be a firewall, VPN, application A, application B, they can be discovered. And if they are discovered, frontier models are likely to find some vulnerability that creates a problem.

In the Zscaler world, we're unique that with Zero Trust, we can hide those applications behind. So we are very busy working with customers hiding their applications, customer after customer. There have been cases a customer had 500 exposed domains, subdomains, applications. And once we go in and we are able to bring that number down to probably 200 in a matter of a few weeks, then it takes more time to go further down. That's number one.

Number two, they are worried about embracing AI agents. They're powerful. They're excited. Every CEO is putting pressure on them, do it and then they say, wait a second. Then they read about Hugging Face. And they even next week Anthropic saying, "Hey, my agent went rogue". Then Meta comes, "yes, mine went rogue too". And then last week, what OpenAI, one of my agents hacked a German site 3 months ago, okay? All this is happening.

So enterprises are worried about it. So for that, they are looking for essentially a solution that can make sure right agent can talk to right agents. There's nobody better suited to solve that problem than Zscaler because we pioneered Zero Trust where a user could only connect with certain application, not on -- not to the network. Firewall vendors connect to the network. We advanced it to Zero Trust for branches, for devices, for cloud workloads. So doing agent was a natural thing for us to do. So those are the two things.

I'll give you one example yesterday. I was in D.C. met a few customers. Two of them are large federal agencies that you would expect in D.C. That's a natural part of it. And both conversations tell me Mythos, they're both using Zscaler, ZIA, ZPA stuff. And now they're looking for help. So now the conversation happens. I talk to the CIO, how we can do this stuff. They are large agency, thousands of one.

And the CIO said, I have 900,000 Internet access points, okay? How long will it take for you to give me an assessment and the risk factor of it? I told that it will be faster than how long it takes you to go through your paperwork for procurement. He said, how about 4 to 6 weeks? I said done. I said, let's get started. We'll figure out. And then our real opportunity comes actually to block all those things. So one of the opportunity motion we have is to show them what the risk is as a complementary service, then the opportunity comes to us. So it's sizable, and we are focused on leveraging it.

Jay, I'm going to take a little bit of a detour and talk a little bit about the historical context a few things, so if you can humor me. You've had the privilege and the benefit of actually witnessing and participating in disruption in prior computing cycles, right? So you saw the rise of the cloud secure web gateway coming. Cloud was ascendant and we were moving away from the client-server model. I think you're famous -- and you've been famously talking about that the castle architecture shouldn't exist. You stack the DVD players, so I think we're familiar with that.

So you lived through that computing shift. So cyber attacks and the rise in volume and velocity and variety and sophistication of cyber attacks was an issue at that point in time, Target, COVID era, pandemic era, there was an impetus for security modernization. And now we're here with the AI native era, right? So cyber attacks have always been a problem. They've always gotten worse. But in your opinion, what is different, more pernicious, more consequential in this period of time that doesn't compare to these past periods?

So the number of attacks and the nature of the impact of attacks has gone up what I may call incrementally. What Mythos models have shown, it's a step function change. It's a big change. Let me give you some data points. The number of vulnerabilities that pre-frontier model companies who do vulnerability management are able to discover, if it is x, frontier models are discovering 10x. That -- so it's such a big.

For example, before Mythos, we would go and tell customers, you should hide your attack surface with Zscaler. It's a good thing. And they'll say, yes, it's a good thing, but it's not -- I have a lot of other things to do. It's this kind of priority, but not the highest priority. Now if something can be discovered, there's a pretty good chance that the frontier model will find vulnerabilities that can be exploited.

So the risk has gone up significantly. These assets have to be done. So that's what makes these people nervous. So the biggest change -- one of the big attacks that happened, what, 10, 15 years ago was Target.com. That was big. It shook things up, but nothing like this. This is -- you're seeing things every day. I mean we were part of the Project Glasswing from day 1, early March. When we tested our own software, we found quite a few vulnerabilities.

I changed priorities in my team and say, your #1 priority is fixing it and the feature delivery can slip a bit, that's okay. So I think that is very important. The good thing is a cloud-native company like Zscaler can fix those things on its own. If you're selling lots of firewalls, then we have to reach the customer to figure out, upgrade all the kind of stuff. But this is the biggest step function change in cyber world.

So just bringing it back to your opportunities -- your prosecutable opportunities and your capturable opportunities. I think one of the things that's been confounding for investors is that the environment has never been more dangerous. I mean you just obviously gave us some very examples of why the market factors and the exogenous variables are absolutely in your favor.

So the confounding element, I think, for a lot of investors has been, well, where is the budgetary explosion, so to speak, right? You would have assumed that the cadence of budgetary growth would have accelerated immediately, right? And you just mentioned to me that the federal agency has a 4- to 6-week time period, right? You also mentioned that in your CISO conversations over the last 6 months, post Mythos, there was a task force, right?

So it's almost like instead of seeing a budgetary explosion that you've been able to capture, there's maybe been a little bit of a pause from an organizational standpoint as to there is such a tremendous amount of change. You want to be deliberate and thoughtful around it. Would you disagree with that characterization, number one? And then number two, how can we -- how can you express comfort and confidence that actually the budgetary explosion, so to speak, is on the come because budgets are only going to need to rise to the occasion of the cybersecurity and cyber attack environment that we're in?

Yes, good comprehensive question. So first of all, I would say the urgency to handle security issues has gone up big time. That's there. Now I will contrast the urgency during COVID versus today. This question has been asked to me. COVID was a catalyst. The difference between the two was the following. When you were sent home on Friday, whatever March 14, 2020, on Monday, you couldn't work unless you had a solution. So we were inundated with calls. We are literally working day and night to turn on Zscaler to get things moving. So it happened very rapidly.

On Mythos side, the risk factor is viewed as far, far bigger. But now they are trying to understand the risk, what, how, where, prioritization. Those things have played a role. So you did not see kind of a silver bullet saying, suddenly, the budgets are jumping up. But the urgency is going up. Now urgency is in a couple of areas. One is these models, what can I do about protecting it? And then the other pressures CEOs are putting on CIOs and CISOs embrace AI safely in a much faster fashion. Now they are trying to figure out what do we do.

Now embracing AI is happening, but bigger rollouts are still slowed down because of cyber. Now when I go and talk about AI security or security of AI to customers, which is every conversation, they are looking for, can you give me a comprehensive solution rather than 5 point products? So that's favoring us. We built about -- we built 4 products internally. We acquired 2 products through acquisitions. So it's helping us there.

The other thing that's helping us is they're seeing there are 1,000 probably security for AI companies out there. The message from CIOs is, I don't want one more point product. I want integration that's helping us in that space. But also, I am also seeing that even though we have done hundreds of deals in the past 12 months for AI security, the deal size has been smaller.

They're kind of say, I want to understand -- and before I do multimillion dollar deals. So size deal are smaller, urgency is there. It's also driving need for Zero Trust because either you hide your applications behind or if you think somebody got in, you need Zero Trust to reduce the blast radius. So branch is infected. It doesn't go beyond the branch. So it's helping data security as well.

By the way, the deals we closed in Q4 for AI security, 70% of the deals also had data security as a part of that. A number of our Zero Trust deals did get accelerated because of the sense of urgency here. So we've seen an impact. I think we expect to see more impact in fiscal '27.

This is a good segue into the next question I wanted to ask you is just around you've framed the business by virtue of pillars, let's just say loose pillars, right? Zero Trust everywhere, AI security and data security.

I'm curious what is now -- has the tip of the spear for you changed? Because clearly, in the way you've described it, there is a lot of coincident value in having a Zero Trust architecture and portfolio of products from you on the Zero Trust side with a layer of data security, and we'll certainly get into more of the data security implications and what that means for your business.

But generally speaking, you framed the business in those 3 particular pillars. I'm curious as it stands today, as you think about fiscal '27, as those budgets open and the urgency is realized and you're monetizing it, where is it -- where should investors expect to see the outcome, the positive outcomes of that budgetary allocation? In what domain would you expect that to show up most?

Yes. So great question. So first of all, those 3 pillars you mentioned, we just added a fourth pillar for Agentic SecOps with the launch. It's a new area. It will take some time, but it's also a big opportunity for us. Where is that traction coming from, right? So I have been surprised by the speed at which the security for AI products have taken off. I told you a year ago, we had basically one product not a whole lot of stuff.

As many products...

As many products. And also, we disclosed after Q3 was our last 12 months bookings for security for AI has exceeded $100 million. And in Q4, we all saw that sequentially, our bookings went up 50%. So we're seeing good, very good traction. I expect some serious traction in this area. Though we haven't quite factored a meaningful upside of security AI in our guidance. We just want to be prudent. We just -- unless we prove we get comfortable, we don't really want to raise expectations.

Take Agentic SecOps, too. We are working with a number of customers in early stage. We haven't factored a whole lot of that in because we haven't really proven ourselves. So the big growth areas, AI security starting from a small base, but it's growing rapidly. AI is impacting data security that is helping. And Zero Trust everywhere is the foundational piece if you really need to be secure. There's nothing more fundamental than Zero Trust everywhere.

We're seeing our branch deals going up. Zero Trust everywhere, number of customers has moved up from 300 a year ago to 950 this year. That's pretty remarkable. We're doing some very large deals on branch. I'm excited with this area, very much so because I've always believed that the network, call it disruption, MPLS, we did that very nicely. People used to say, you're crazy, if you say MPLS is going to go away, it can never go away. right?

Now Zero Trust branch is disrupting SD-WAN. Even though Gartner talks about SASE and SD-WAN, we kind of on a different TAM. We are in SASE without SD-WAN, Zero Trust branch and all that kind of stuff. We're doing some very big deals in this space. We're also doing deals where Zero Trust branches bought without having to buy traditional ZIA ZPA because they want to do segmentation.

So one thing new this year for us for growth is we have now multiple landing solutions. And even though we had some before, we never focused on it. Now we're realizing that you go to a customer, if you go with one solution and the timing may not be right for it. But if you have 4 offerings, they're more likely that you're going to say, I need this.

So for example, AI security, a lot of deals we're doing are new logo deals. Data security, a number of those products are new logo products. Zero Trust branch, when I do plant segmentation or all, I don't need this solution. So I have multiple solutions that are helping us expand our platform. I mean, we shared about 785 customers now with $1 million-plus ARR, $5 million ARR number has been going up as well.

We did a record number of million dollar deals last quarter. We created a record pipeline last quarter. Our productivity in Q4 was the highest in the past some 4 years and productivity of '26 has been very good as well. So that sounds bullish about '27.

There is kind of a running theme that you're alluding to around this resurgence in Zero Trust principles, the foundational underpinning to build your agentic workflows and agentic applications. So as part and parcel to that, historically, predominantly Zscaler's business model was sort of a user-based model, right? We've always been acting as a switchboard between the user and the end application, right? I know increasingly, you've been moving towards becoming the communication switchboard for agents, right?

And branches and workloads, yes.

So as kind of the end user units, let's just call it that, move away from humans to machines, to agents, how are you levered financially to those positive trends? Because I think we've all heard the stats here that an agent-to-human ratio in the next 3 to 5 years could be 50x, 90x, 100x, right? Sky, maybe the limit on that. So how is the business model benefiting from that today? And where do you expect that to be? And should that actually show up in Zero Trust momentum? Or will that show up in the Agentic SecOps pillar that you just launched? How would this -- what is the culmination of this growth of end user units?

So it'll show up in both, but the starting point will be Zero Trust Exchange. Then the logs from there flow into Agentic. Agentic SecOps become the secondary part of it. Look, the more agentic communication happens, there'll be more need for policy enforcement who talks to who. That's where our exchange or communication hub comes in.

And some people have made a case that, oh, this exchange, this policy should run on the endpoint or here and here. It can't run on one entity because agents will be on the endpoint, they will be sitting on my mobile phone. They will be coming from Snowflake, ServiceNow and hyperscalers. So you need almost like a phone switchboard in the middle that can connect any party to any party. So that's part number one.

And with that, we will scale. If you look at it, the reason we believe we have the right to win in Agentic communication more than anybody else is because we have the core competency. We built the exchange for users, to applications, then took it to branches through devices, to cloud workloads, to Agentic Exchange. Now literally to build what we have to build, somebody who doesn't have this background has to set up infrastructure around the globe. Then they need the policy engines, logging reporting. literally about 70% of what we needed for Agentic Exchange, we already had it. 30% is what we have been building.

What is that 30%? What's different? Well, MCP gateways, A2A gateways are starting communication point. You talk -- you communicate our prompts, ability for your proxy to extract prompt, analyze it, understand the intent, understanding potential things had to be done. So guardrails needed to be built. Intent had to be understood. Identity, we have always taken identity from whoever provided it.

Agentic identity is a little bit more cumbersome, a little hard. A user identity is kind of static, you got it. Agent can change on the fly. Is identity for a given session, how dynamic is it? What all can they do? If they're launching 5 subagents, do my permissions go to my agent and what all do they go? So there are a little bit of authorization pieces. Those are the pieces we built.

We launched our exchange, in June at Zenith Live, and it's in limited availability, working with a bunch of customers to get all the things out. But that's the biggest opportunity with biggest barrier to entry. Then logs flow into SecOps. SecOps is there to see if you didn't catch something when it's happening, it's after the fact, it's needed.

Then with our logs, we don't have to spend time to move it to a data lake and get the results back. Today, it can take days or weeks to find something and take an action on it. With Zscaler, where we can find things in minutes because we have the first-party data, take an action to our exchange to block up and do a policy enforcement. That's really what makes it exciting for us.

Jay, you brought up identity, so I have to ask you this question. I think there is a debate I certainly have with a lot of folks around, well, what's the right way to think about the identity architecture that you wrap around an agent? Clearly, it's very different from a human user like you characterized. But ultimately, do you believe this is a privilege-oriented problem? Or is it a governance-oriented problem? And technically and even conceptually, how do you solve for that agent's level of permissions, right?

So identity can fall in 3 buckets for your simple understanding. Base is who are you? That comes from traditionally for users, Okta, Microsoft, Ping or the Vault. Then the next level is what all can you access. Within SAP, you can access this. That's called authorization and governance that comes from SailPoint of the world.

The third piece is what special privileges can you have? That comes from privileged access management kind of company. This is BeyondTrust and CyberArk of the world. These are there.

Now start with identity. You need to start with identity of the agent. In my view, as we work with all the hyperscalers and Snowflakes and Salesforce of the world, when they create agent, they create identity. So they're going to go to an independent company and say, use my identity for this agent, gets very hard. That's why our decision is take identity from the party that creates agent. Now we can add authorization policies on it, figure out intent and the like. That's the right way to do.

A lot of companies who are trying to say, I'm the identity company, they're taking machine identity, many times certificate-based stuff, which is good for servers. It's not good for agents. So I personally don't think that's the right approach. We will augment identity authorizations on top of identity we get from the agent providers.

And is that identity intellectual property going to be your secret sauce? Or is that something that you'd be willing to partner with traditional IDP and identity.

Identity is going to come from identity providers. We add value to figure out whether the policy should be enforced or not, checking the intent, checking authorization risk associated with that. That's the value we add in our exchange. Exchange is not a simple thing anymore. It's more sophisticated. And that's our IP.

Jay, when we talked about, again, the multiple landing spots and landing zones for the customers, you've got the go-to-market motion, but something new that you've also introduced over the span of this last year is Z-Flex. So just as a procurement model and the procurement vehicle to drive more awareness of the full portfolio. Can you give us the top three learnings that you've gained from launching Z-Flex? And any metrics that you're proud of in basically having this be a very important sales conversation tool. And what should we expect from Z-Flex as we think about fiscal '27?

As the name implies, Z-Flex, it's a flexible mechanism for our customers to acquire our products. So Z-Flex is not a new product, new packaging. Z-Flex says I want to reduce friction for customers to be able to do larger deals with us, align with us and it becomes a win-win.

So what are we offering with that? Number one, customer used to look at, you got 6 things, I'm interested from data security. I want to test A, then B, and then C, will decide what I want, what I don't want. And so we give flexibility to all modules. You don't have to keep on testing the stuff. You can keep on moving faster. That's number one.

Number two, create a rate card so the customer can buy additional product without going through procurement cycles. Anytime you go through a procurement cycle, it just drags things on. Number three, with the flexibility, they also want many times, though not always, some ramp. I got 6 things. I would have bought 3 otherwise. Now I can't roll out 6 at the same time, give me 3 months, 6 months, 9 months ramp time. So ramp becomes part of it.

And with that, they're also willing to do longer deals. So more and more 4-year, 5-year deals. So it's a win-win thing. So we are seeing deal size going up. Customers who have done Z-Flex with us in the past year, it's in less than a year. Their upsell has been 30% higher than customers who haven't done with us. I think it's great for customers, great for us.

Now we need to take it to the next level. We started less than 4 quarters ago. Just first quarters are testing the water to see what limited number of things, then we opened up more and more. And this year, we want to take it to a broader level because it's a good thing for customers.

I did want to spend a little bit of airtime talking about the big launch that you had today of the Agentic SOC and Agentic SecOps. This is something that you have telegraphed on the back of the Red Canary acquisition and what your anticipation was after a full year of integration. So can you just walk us through the culmination of -- you did the Red Canary deal last year. How has that now been fully folded in and integrated and melded into Zscaler proper?

Red Canary was done to make sure we can accelerate the delivery of our Agentic SecOps solution. since they had expertise in SecOps management as a managed service, they have playbooks, they have built some very good agents. We took those SecOps agent, embedded them into our solution. That's what brought together. That's what the launch is about.

And so it was never about us fully saying, I want to be an MDR vendor. We want MDR management as an auction managed service. But SecOps is a great solution. Our enterprise customers want it. So it's a product they can buy and they can also buy a managed service if they choose to do so. So from a revenue point of view, actually, Red Canary has been a headwind for us rather than a tailwind because MDR stand-alone companies traditionally have had elevated churn rates, and we saw the same thing here too.

And there are also customers on lower end. You can't -- the smaller customers mean you can't really touch them that often that generally ends up having higher churn rate. The success of SecOps for me is how much revenue do we grow to our traditional enterprise customer base. There are big deals to be done in this area. We expect that. And we'll plan to bring the Red Canary customers over from the current solution to the new solution because we aren't going to keep two solutions. And we haven't fully tested how much uptake that's going to happen.

So we factored some of that as a headwind from Red Canary. That's why we're not expecting much from the Red Canary side of it, but we are expecting this to really build momentum for our overall SecOps solution to our traditional customers.

Just from a competitive standpoint, Jay, there has been a huge battle to win the hearts and minds and the budgets of the SOC because we are in this generational modernization period. So what would you articulate is kind of your de facto advantage in the Agentic SecOps, security analytics, stack modernization opportunity? And why do you have that must-have angle relative to some of your peers who may have had a little bit more of a head start in this domain?

Some folks had a little bit of head start in the domain, but all the head start was in pre-Agentic era. True Agentic technology actually happened in the past 18 months, not even 2 years ago. A lot of stuff before that was design, mimicking human beings, doing a little more automation, doing even faster.

We had the advantage of doing it in the past 12 to 18 months in a new way designed for agents, not for human beings. That's number one because agents will do more so what. That's number one from a technology point of view. Number two, from a technology point of view is closed-loop system. You send it to some system, they take a week or x days to discover it.

And all those remediation has to be done somewhere in an in-line system, either on the endpoint or in-line system like Zscaler. We being able to offer both, not having to move the data to another lake are able to do some of this stuff in minutes. That's advantage, especially more so in today's world when you find an exploit and it gets exploited pretty quickly.

And third, big thing is data. We really have some of the best data. People make a case for whose logs are better logs. Well, in line is fundamentally important. Imagine you need to figure out how do I keep my country safe and you know logs about who is moving in and out of your country. The best place you know who is coming in and out. And the EDR is important because EDR sees what's on the endpoint, endpoint become a point of exploitation. We have both. We correlate both. So with that telemetry, we think there are right to wins with this thing. Lots of customers are asking us, you create all this telemetry. Why don't you give me more value out it rather than I having to pass it to someone else.

Jay, my last question for you is in September 2027, when we're going to have you back at the conference, what will have been the two things that will have happened that would have surprised you to the upside? So instead of doing 17.4% ARR growth in fiscal '27, you're going to do 25% ARR growth. My number is not...

Only 25%? Cool.

The two things that would really surprise you to the upside in a look back from a year from now?

So number one, security for AI growth, okay? I'm very bullish. But again, my bullishness and giving you the guidance need to be kept in the right place, okay? But I'm seeing it. Number two, Zero Trust everywhere would be the other thing. But in that piece, I'm excited and curious to see how the Agentic exchange takes off. Now some of this stuff will take -- Agentic will take some time because there's so many unknown pieces that are being built every day. But I would like to be sitting here next and say, I've proven enough to see Agentic stuff taking off because that opportunity is boundless the barrier to entry to do what we are doing in this space is very hard.

Thank you, Jay. I always walk away much smarter after our conversation. So I appreciate your time. Thank you so much, everyone.

Thank you.

Good morning. Thanks so much for joining us at the Zscaler session. Delighted to have Adam Geller, Head of Product; and Kevin Rubin, CFO, on stage with us. Really grateful to you guys for taking the time.

Thank you.

So Adam, I remember the original days of the Zscaler IPO. And one of the most foundational things that Jay and the team drove home was this idea of a true proxy architecture and how taking shortcuts to accomplish what looks like a proxy architecture from the outside in is not the same thing as delivering a true proxy architecture.

So my question for you is, we're now in a world where there is a decent intellectual argument to be made that agents should communicate on a proxy architecture. As a leadership team and as Chief Product Officer, how do you take a vision that was built for SASE and Zero Trust and extend it to an agentic proxy architecture?

So I think the -- so -- and I wasn't there during the IPO, but I was at companies that were competing with Zscaler without proxy architectures. So certainly know the space well. I think that the argument on all of it when we think from a product perspective is what's right fit for purpose of what problem you're trying to solve. So you have to come back to some sort of first principle there.

And the reason why a proxy architecture is powerful in cybersecurity specifically is because when you get in the path of something, you have the chance to actually see and inspect because the whole point of doing cybersecurity is you're looking for either threats coming in or you're trying to understand where data is moving and should it be moving or not. So just connecting things serves a purpose.

That allows you to get from point A to point B. But the whole reason why you have that concept of a proxy is it's an inspection point and it's a chance to at least see and observe, you can observe, take notes and take no action or you have the chance to enforce policy. And so that architecture time and time again comes back up as a staple of cybersecurity products and technology because it works and it's extremely powerful.

Now, it's -- like any security piece, it's friction in a process. So you have to be careful how much friction you introduce into a process. And with agents, right, agents are not operating in silos. They're operating by connecting to other data sources. They're connecting by working with other agents. So you have a whole interconnection to consider and that is ripe for can you -- where you apply security into that?

What do you do at the edge? What do you do in that connectivity path? And so that's why we think that proxy architecture is very powerful. The challenge will be, and the key is in, how do you apply it and how do you -- what's the same and then what's different when that traffic is agent traffic versus any other kind of traffic.

Let's stay on this thread. So I think we're all familiar with classic web proxy, where if I go out to the internet, the Goldman Sachs policy Zscaler enforcer will be in the middle, and it will orchestrate or proxy the connection. With agents, the question we get on agents is both, look, they are ephemeral, they spin up and down much more quickly.

And fundamentally, there are things about how an agent runs a session that make it difficult to properly, to your point on earlier, enforce and observe, it becomes different. Maybe just give us the layman's version of how do you take the tooling that was originally built for web proxy and extend it to agentic proxy?

Sure. I'll give you -- I'll give you one very vivid example that I will try to drive that point home. In most communications or most sessions of the human driving it, like I was describing before, you want to understand should I allow something or not? And you basically have milliseconds to do that because anything longer than that when the user gets -- think something is wrong and they go away and they file an IT ticket or they just find a path around it.

So you don't get a long time to decide, is this a good website or is this a bad website or should I allow this transaction or not? And so a lot of, as an example, in in-line DLP, you look for a very quick pattern match. You have that sensitive content, my policy says sensitive content isn't allowed out, so it's not allowed, right? And you make that real decision.

In AI, and this can be both human working with, let's say, a generative AI chat interface or an agent, the iteration, the back-and-forth conversation is actually how that works. So the intent of a one-line request that mentions cookie, you could say that's -- talk about your cookie recipe, right? If it's a back and forth over 6 different turns, it might be about a session cookie that's holding an authentication token for a particular transaction and that might be much more sensitive information that's being sent out.

So you actually have to look at a -- using AI terms, a context window, right, or understand multi-turns of a conversation. So the way you inspect is different and what you're looking for also starts to change depending on what you're trying to protect. In more classic world, it's -- you're looking for vulnerabilities or is this malware or is this an exploit coming down?

Or is my data leaking? In the AI world, you might be looking at is someone trying to the -- is someone trying to poison my model, is someone trying to get it to do something it wasn't supposed to do. There's no vulnerability there. Like, there's no signature for I exploited a vulnerability. That's the way how a lot of cyber works. And when you're proxying AI communications, you need different kinds of inspection engines for that traffic.

Maybe the other way to talk about this is through a little bit of a customer journey, and Kevin, we'll bring you into the conversation. Talk to us a little bit about when you meet with your large customers in the last 5, 6 months, how is the conversation between you -- between Zscaler and the customer changing that addresses some of these new AI risks more comprehensively? And maybe the answer is Zero Trust architecture. But give us a flavor for what those conversations look like on the ground.

Sorry, she said Kevin. I thought it was coming to you.

No, no, either way.

No, no, it's okay. Look, I think the conversations have obviously accelerated and have gotten much more intense post Mythos. And companies deal with the reality that they have vulnerabilities and exposures, many of which they didn't even know they had before, right? So their exposure, their threats are far greater today than it was 6 months ago. And the questions are, how do I deal with that, right? What do I do?

And I think the reality is that patching alone is not going to solve the problem. You've got old equipment that in order -- may not even be in service any longer, you can't patch. You may have old operating systems that you'd have to upgrade before you patch. I mean, the complicated environment of being able to address this with simple traditional methods is becoming harder and harder and the exposure and the time to a breach is becoming less and less.

And so the conversations are, what do I do? And our answer is very, very simple. You need to hide your applications, so completely reduce your attack surface and eliminate the ability of any potential breach to be able to move laterally. And so a lot of the conversations are around how do we address the risk today and how do we help you going forward, be able to secure your environment given how everything is changing so quickly.

Your point on patching doesn't solve the problem. I would love to use it as an opportunity to ask you about frontier models, and Adam, this one is probably for you. What do you think the role is of a frontier model in security? Where do you think the swim lanes ultimately converge versus stay separate?

Great question, TBD, just like it's been in -- and it will evolve, just like it's been evolving with cloud for many years, right? The evolution of what should you expect your hyperscaler to do for you. And then when I was in cyber in that time, in the early 2010s, right, there was that initial belief was, well, this will all be just baked into the hyperscaler now. So I don't need third-party products anymore.

And the investor conversations were all about, so when does this degrade the cyber business, because I was working in another cybersecurity company, and it ultimately didn't play out that way, right? There is embedded functionality in the hyperscalers. And then the third-party elite and very focused cybersecurity companies have found a way to continue to drive value for their customers, right, because that is their core and only focus. And so I think we're going to go through a similar pace with that or evolution with that on the frontier AI models, right?

Every -- as an example, every AI model has the concept of guardrails, right? They also have the concept of running without guardrails. And there are baselines. And then there are what other guardrails, what third-party products like Zscaler makes one AI Guard, that is a guardrail solution for making sure that your model does not go awry, right, either shares too much information or gets compromised in some fashion.

And so I think you'll see baseline level of security capabilities will be built into all the foundational model solutions, and then you'll see those natural kinds of extensions. At least maybe 10, 12 months ago, you also saw this angle of, will the frontier model companies just build cybersecurity products outright? And that's -- it's always a risk. I mean, it hasn't quite played out as scary as it was set up 10 months ago.

When I say scary, meaning just there's no need for cyber anymore because you can just code -- everyone can code it themselves on a weekend. That hasn't proven to be true yet. We haven't seen that from any of our customers. Do we see people using it? Absolutely. But there's a learning curve there, and there's an investment curve just like using any powerful technology, and there's a cost-benefit of do I want to have an entire cybersecurity product practice in my business?

Or is my business something totally different than that, and I'm going to leverage an enterprise product? So we -- so I think that from the model companies, there will be the embedded pieces that they do. They may find certain unique areas where they say, we just think we can be really good at this, and we're going to introduce a competitive cyber product.

Every one of the model companies, though, are competing with -- are partnering with the big cybersecurity players, too. So again, just like in the hyperscaler space, it feels like there's going to be room for most. I think the biggest angle or biggest opportunity that I see is in software development, right? And it's not a particular place that Zscaler plays.

But if you are writing -- if you're more and more using agent frameworks to write your software, the frontier models, I believe, would argue that the ultimate way in which you get out of this, continue this, how do I patch vulnerabilities show up model is if you just -- some level completely give it to AI and say, AI is going to basically do your entire -- almost your entire software development lifecycle because it's the only way you can keep up with the development and then the management and the continuous patching, testing, redeployment of your code.

I think that is a very high bar for most organizations to get comfortable with, certainly large enterprises that we sell to, allowing that level of autonomous changing happening in their mission-critical infrastructure. The human discomfort with that has been a challenge with automation and cybersecurity for many years.

Like I can trust -- can I trust that decision to be made if it's going to take down the trading desk? Or am I willing to say, no, I'm not going to trust that. It can let me know if something is bad and I'll deal with it in off hours. So I think that's probably the area where I would expect to see more and more -- and where I think the frontier models are very well suited given what they do.

What about for -- and this actually comes back to the R&D part of the conversation as well. What about in terms of your own applications of AI within your organization? Any thoughts on how you can apply something like an open-source model or even a frontier model to your own rich data sets in-house to have an evolution in data plus model to get to something greater?

Yes. I mean, in terms of just AI internally, I mean, I think we're going along the same journey that many companies are. We have a very strong initiative internally to continue to use and deploy AI to improve basically fundamental productivity. We think about it as employees plus AI, not humans or AI.

We do think that they're -- the pairing of those 2 are kind of where the powerful combination comes into mix. Some of the early areas where we've had more success, like many companies, customer support, ticket deflection that has been a very significant kind of continuous evolving area of success in the go-to-market space, we've been continuing to expand use of AI.

And then in back office, we continue to expand. So I think there's a lot of opportunity going forward. I think we're just scratching the surface as to what's available. In terms of how we use AI and data, maybe, Adam, do you want to talk about the launch this week?

Yes. I'm just going to finish one quick point internally, and then I'll talk more externally with customer data. But I think that the -- Kevin is right, all of the different functions are looking at, well, how do I leverage AI to change how I operate, right? And on the engineering side, it's definitely changed how we're developing software. It has not only changed how it deploys because that gets back to the humans and the customer side of it. Sorry about this, need a quick drink. Tail end of a cold and talking for 45 minutes straight.

I got it. Yeah.

It's not great...

Give me a wink and we'll make Kevin answer.

Make Kevin answer more questions. Maybe I'll pivot quickly. So we do have our Agentic SOC launch today. And that is all about taking the 750 billion signals a day that Zscaler sees and running it against an agent framework for helping customers detect, investigate and respond to security incidents. And that's something that we're really excited about.

That is not a human readable problem, and it was ripe for agentic transformations, which is true for everyone who's operating in the security operations world right now. The difference for us is that you could maybe joke and say it's a late mover advantage because we know the space is not new. We didn't start with, let's just build a giant data lake to collect it and then do a little bit of normalization and then create alerts for humans to look at, which is what the traditional SOC models have been built around.

They're all evolving. We started with how do we build a model that is -- how do we build an approach that is set for agent-first interaction. And some of that we got through our Red Canary acquisition, and we've been bringing into this, and others are new agents that we've been building, but our starting point with our Agentic SOC is effectively a family of dozens of agents that are doing that detection, investigation and response process. So we're excited about that.

One of the more common pushbacks we hear on competition in agentic SOC is, well, there are all of these other companies that have really strong footholds in endpoint and SIEM and it's a much more natural progression for them to own the agentic SOC than it is for Zscaler with the network background. Maybe push back against that a little bit and talk about why your architecture is the right one versus a CrowdStrike or a Palo Alto.

Sure. It's -- having been in the SIEM and SOC space for a long time before Zscaler too, endpoint data is certainly very valuable. That argument immediately gets -- the pushback to that argument immediately goes to, look at all the agentic SOC companies that are in the start-up world right now, none of them have an endpoint presence because they're tiny companies, yet they're all claiming they can do agentic SOC.

So you do need access to certain data points. And it is true that owning a vantage point is powerful. I think the guess what or the surprise is, well, Zscaler actually owns 2 vantage points. So we certainly own the network vantage point, but we also have 60 million endpoints. So our endpoint presence is probably bigger than Palo Alto Networks endpoint, not bigger than CrowdStrike, but it's bigger than Palo's.

And it's not an endpoint for just connectivity, right? It is an endpoint that does DLP, that does policy inspection, that does user performance, that does software inventory. So we have all that data set, and we were never bringing it together other than for those micro use cases like, okay, I have an endpoint, I'll use it for endpoint DLP. Now all of those signals at the endpoint and in the network are coming together.

So the kinds of incidents that we're seeing our Agentic SOC be able to detect and investigate certainly leverage endpoint data. They certainly leverage EDR data from partners like CrowdStrike and every other -- and Microsoft Defender and SentinelOne and others, but it's just one of many pieces. And what we think is that, especially in a world where the threats are increasingly going to be less, here's a piece of malware that someone is trying to download.

Here's a behavioral pattern instead of what happened on this endpoint, what machines did they access? Where did they try to go on the internet? Well, Zscaler sees all of that telemetry. So we are -- we have equal, if not more right to bring that picture together for a customer than someone who's just looking at one of those pieces. So -- and that's what will play out, and we'll see how that plays. But we've run a threat hunting service for a number of years on a small scale that's been doing just that for customers.

And what we find is that while the red alert sometimes may come from an endpoint trigger, the first thing people look at when we say, oh, what do I do to investigate this person or an agent is, where did they go? What were they doing? And they immediately then go look at the Zscaler logs, right? So that's when it comes in. And so we have all that together, and we believe that our agents can actually connect that and then we have our human expertise on top of it.

That all makes sense. So Kevin, there were a number of product cycles just in the last 20 minutes that we've hit on. I want to map it a little bit into how the sausage was made as it pertains to FY '27 guidance. You've also acknowledged that there have been moving pieces to the sales organization. So maybe let's take a step back.

How did you think about normalizing, perhaps that's the right word, for some of the changes that you're making in the sales team relative to some of the excitement that Adam and your customers are talking about, about the various product cycles within Zscaler that are more than just core networking or core SASE.

And I think we're in a really exciting, interesting time, right? We've got momentum building from an AI perspective. We've got -- I mean, Adam kind of alluded to it, but I want to make the point a little bit more fine. We now have opportunities with these new products to land customers in ways we didn't before, right? So if you look at where Zscaler was just a few years ago, we were largely going in and starting with prospects in terms of user transformation. Today, we have an opportunity.

We can still obviously do user transformation, but we have an opportunity to go in and have a conversation around a manufacturing facility, a branch. We have an opportunity to go in and talk about AI. We have an opportunity to talk about when you think about exposures as we were talking about with AI going in and saying, let us help you solve your inability to patch fast enough and hide your applications so that they cannot be publicly identified. We have an opportunity to go in and talk about the Agentic SOC.

So the ability for us to go in and have very different conversations with customers and prospects is very different as we think about today versus just a few years ago. Having said that, a lot of these are very new, right? When we talk about the Zero Trust Exchange for agents as an example, right, that's an early access. It's not actually generally available. A lot of our AI capabilities have been released in just the last 6 months. And so a lot of this is new building a momentum.

We're very excited with the traction that we're getting, but it's also new, and we have to balance that. The piece of the business that has been very consistent and predictable has been our upsell, right? We've been able to consistently deliver NRR of 115% for the last 4 quarters by way of example. And we have confidence in our ability to continue upselling. One of the points we didn't address yet, and I'll just touch on it briefly, is Z-Flex, right?

We introduced Z-Flex about a year ago. It's our flexible pricing and packaging offer for customers that are looking to make longer-term larger commitments. And in those customer contracts, we've seen an uplift of 30% against, kind of, the before and after ARR picture in that environment. And so there are a lot of, I think, real positive signals as we go into the year. We did obviously raise guidance in the last quarter.

We also know that we have an opportunity to do better in new logos. And so we've got specific initiatives as we think about fiscal '27 go-to-market around new logos. We have about 25% of the enterprise customers that we target today, about 4,600, against the population of a target audience of about 20,000 customers. And so that's really where we see opportunity as well.

Give us an example on how you can change the new logo success.

So 2 things I can point out. First of all, we are investing more in sales reps that are solely focused on enterprise land, right? So going after pure prospecting territories as opposed to a blend of those. Along with that, every year, you kind of make changes in how you think about monetizing and addressing the sellers to incentivize them to do that.

Another example we mentioned was an agreement with Carahsoft, who is going to take a stronger role down market for us where we don't devote a lot of resources, but there's a lot of opportunities. So we're really looking at ways in which we can continue to enhance it. And then back to the ability to land with multiple products, making sure we have clear enablement and sales plays that assist reps as they're having those conversations and making that front and center.

As an analyst, we sometimes try to predict the inflections in sales and marketing product cycles, which, as you can imagine, a little bit of fool's errand from the outside in. If you had to help us out here, talk to us about when you think we might start to see a more meaningful change in productivity? Ramped reps or -- however you think of it from a cohort standpoint?

So Q4, as I mentioned on the call, was the highest quarterly productivity within the sales organization that we've seen. So very proud of the continued improvements in sales productivity in the business. And fiscal '26 on an annual basis was the highest productivity since 2022, so about 4 years. So we are seeing significant productivity improvement that we do carry into the fiscal year.

So very pleased with where we are and the improvements we've made. We do have 2 leader transitions that we are going through at the moment. One was a vertical leader where we promoted an internal candidate who's now responsible for that piece of the business. The other was a geo leader where we have -- we've identified a candidate, that candidate has accepted, we just need to go through the onboarding process.

It's outside the U.S., so it takes a little bit longer to land. So I think we are seeing all of the right signals in terms of kind of going forward. It's just that as we think about the products, how they land and how they build momentum, the pacing of that is different, right?

So the Agentic SOC, as an example, we think that that's probably a bigger '28 event as we kind of roll that out into market today and start to build momentum. Zero Trust Exchange for agents, again, is probably a back half of the year phenomenon. And so as these things stack upon themselves, we'll start to really continue to kind of build on the momentum we saw last year.

We'd be remiss not to spend a couple of minutes on ZIA and ZPA. There are 2 schools of thought. There's the bear school that, look, the best part of this product cycle is behind you. You've already gotten really good penetration in enterprise. There's a bunch of new competitions.

Then there is, I think, what is the more balanced view, which you have subscribed to, which is we might actually have a renaissance on SASE with things like sovereign SASE and network traffic going up. Maybe just flesh out some of the pieces of the bull case for the core Zero Trust business, ZIA, ZPA into 2027.

Well, I'll start to give Adam some more time to address his voice. I think it goes back to a couple of things that we've talked about. So first of all, within our target market, we only have 25% penetration in those customers, which means we have 75% of customers who have yet to adopt and deploy Zscaler. So that, to me, is a large opportunity for us that we see as an opportunity.

Within our existing customers, we generally will -- if we're going in for user transformation, we'll generally consume the user population, but that doesn't mean that it's a one-for-one with ZPA. So I still think there is significant opportunity within the existing customer base around ZPA, ZDX and then some of the other components.

And we touched on it a little bit earlier, but as it relates to AI and how do you get your hands around potential threats in, I think Adam used the terminology line speed, but moving at agent speed, it is going to renew interest in Zero Trust and how do you build a security environment that can operate at speeds that humans can't. And I think that, that will turn the conversation back to the Zero Trust architecture that we've adopted and deployed.

Maybe one thing I'll just add to that and tie back to your earlier question about how Zscaler is using AI internally, right? So even -- and this is going to be true at every company, I believe, or many companies. We have people this past year who built applications who are not developers. Someone on my team who built an entire operating system for how product management works, and it's an incredible application. It ties together Jira, Slack, Salesforce, and it was built by 3 people.

That application is becoming critical for my team to run. It needs to be protected. It is behind ZPA. That's a new application in our company that we didn't buy, but it absolutely needs to be protected. It needs to -- it has an AI interface to it. I need to make sure that when a salesperson chats with it, they get what's publicly available. They don't get the whole road map. But by the way, the model knows the entire road map that's there, right?

So all of those protections that we've been using, they actually need to now protect that application. So back to that case of will there be a renaissance? Well, if you tie in and say, what is AI going to make easier, the generation of software and code, it's not just to run -- it has to run somewhere and it needs to be protected. So I think certainly, even if you could say user counts might not go up or go up dramatically, the application proliferation is going to be huge, and those all need to be protected.

Fantastic. Kevin, were you about to...

I was just going to say, and if Jay was sitting here with us, kind of calling on his point of view, he would say that the opportunity to secure agent-to-agent communication is, in his mind, significantly larger than the opportunity to secure users. And so that's an important piece.

That's the perfect place to end it. Please join me in thanking Adam and Kevin for their time. Thank you, guys.

Thank you so much.

Good day, and thank you for standing by. Welcome to the Zscaler Fourth Quarter 2026 Earnings Call. [Operator Instructions] Please be advised that today's conference is being recorded. [Operator Instructions]

I would now like to hand the conference over to your speaker today, Kim Watkins, SVP, Investor Relations and Strategic Finance.

Good afternoon, and thank you for joining us today. Welcome to Zscaler's Fourth Quarter Fiscal 2026 Earnings Conference Call. On the call with me today are Jay Chaudhry, Chairman and CEO; and Kevin Rubin, CFO.

Please note that we posted our earnings release, shareholder letter and a supplemental financial schedule to our Investor Relations website. Unless otherwise noted, all numbers we talk about today will be on an adjusted non-GAAP basis. You will find a reconciliation of GAAP to the non-GAAP financial measures in our earnings release.

Before we get started, I'd like to remind you that today's discussion will contain forward-looking statements, including, but not limited to, the company's anticipated future revenue, annual recurring revenue, net new annual recurring revenue, operating margin, gross margin, operating profit, net other income, earnings per share and free cash flow margin, our customer response to our products, our expectations regarding AI and its impact on our business and customers and our market share and market opportunity and our objectives and outlook.

These statements and other comments are not guarantees of future performance, but rather are subject to risk and uncertainty, some of which are beyond our control. These forward-looking statements apply as of today, and you should not rely on them as representing our views in the future. We undertake no obligation to update these statements after this call. For a more complete discussion of the risks and uncertainties, please see our filings with the SEC as well as in today's earnings release.

I also want to inform you that we'll be attending the following conferences this quarter: Citi 2026 Global TMT Conference on September 9; Goldman Sachs Communacopia and Technology Conference on September 9; Wolfe Research TMT Conference 2026 on September 9; and JPMorgan 2026 Software Forum on October 1.

And with that, I'll turn the call over to Jay.

Thank you, Kim. Good afternoon, everyone. We delivered a strong finish to the fiscal year with 25% ARR growth and a non-GAAP operating margin of 24%. We are seeing significant positive trends in net new ARR with growth, excluding Red Canary, accelerating to 17% in Q4. These results reflect increasing market adoption of our Zero Trust platform. AI is quickly becoming the largest tailwind we have ever seen driving demand for our Zero Trust everywhere, data security and security for AI solutions.

Our customers are relying on us to both combat the threats created by Agentic AI and and safely deploy AI agents and models at scale. Since our last earnings call, new AI models have become more powerful, more autonomous and more dangerous from a cybersecurity perspective. The ability of these new frontier AI and open weight models to uncover previously unknown vulnerabilities at a rapid pace has become even more apparent. Organizations simply do not have enough time or resources to fix the unprecedented number of software vulnerabilities that are being discovered.

The time between vulnerability is being discovered and exploited is also shrinking. You cannot win a patching race against AI. This is leaving organizations exposed and likely leads to more breaches. Over the past few months, our team has conducted hundreds of frontier AI risk assessments for global enterprises to help them assess their cyber resilience posture to minimize potential breaches that takeaways have been incredibly revealing. Over 90% of organizations had AI applications, models and other servers exposed to the Internet and more than 1/3 had known exploitable vulnerabilities.

At the same time, the threat landscape has been compounded by the ability of new AI models to power ungoverned autonomous agents. We are now seeing the impact of fully autonomous AI attacks. They have already been multiple high-profile incidents from 3 leading frontier model companies where agents went rogue and took unauthorized actions. This includes the recent Hugging Face incident where a swarm of agents went to extreme lengths to break out of a sandbox training environment, get onto the copay network and move laterally to conduct a sophisticated attack.

This is driving urgency at the highest levels and Zscaler is uniquely equipped to meet the moment. CEOs and boards are looking to us as their trusted partner to address 3 essential business challenges. First, how do we secure our environment when we can't patch security vulnerabilities fast enough? Our Zero Trust Exchange makes applications, AI models and data invisible. And attacker, human or agent, cannot reach what it cannot reach.

Second, how do we minimize the impact of a potential breach? Our Zero Trust Exchange connects users, workloads, branches and agents directly to the applications they need without placing them on the network. This eliminates lateral threat movement containing the impact of a breach.

Third, how do we take advantage of all the benefits of AI without introducing significant new risks? Zscaler has a full portfolio of data security and security for AI solutions that prevents data exfiltration, provides guardrails to prevent abuse or misuse of AI applications and enables secure agentic communication.

Because of these advantages, customers trust us to secure their business-critical environments. We are also differentiated by our scale, operating the world's largest distributed in-line security cloud, processing more than 750 billion transactions per day. This scale provides unmatched high-fidelity telemetry that continuously improves our AI-powered security capabilities.

In the AI era, Zero Trust is now an imperative, and we're not alone in this belief. Anthropic published a white paper in late May, encouraging adoption of a Zero Trust architecture for AI agents, emphasizing the importance of treating every agent like an untrusted entity and making sure it only has access to authorized data and applications.

This is why customers are expanding their investments with us to secure their agentic infrastructure and why we are confident our platform is uniquely equipped to address the risks companies face in this new world.

In addition to protecting companies from the threats created by Agentic AI, we are also enabling organizations to safely deploy AI agents and models. Our AI solutions are key to providing enterprise visibility, governing what data and applications agents can access and what actions they are permitted to perform.

We are often asked why our solutions are needed alongside identity for AI security. While identity solutions answer who is requesting access, our in-line exchange determines what that user or agent should be allowed to do and enforces that policy in real time. Put simply, identity is only the starting point for securing AI. Greater visibility and control is needed for organizations to trust agents accessing sensitive data, interacting with applications and taking action on behalf of users.

Earlier this year, we introduced the industry's most comprehensive security for AI solutions designed for exactly that reason. We are seeing strong proactive inbound interest from both new and existing customers, and we have seen no budget constraints. Security for AI bookings increased more than 50% sequentially in Q4 on top of a strong Q3. Our security for AI solution provides new logo opportunities by offering organizations an integrated way to secure AI use at scale.

At our Zenith Live Conference in June, we unveiled the latest additions to our security for AI lineup, including our Zero Trust Exchange for agents and Endpoint AI Security, both of which we expect to scale in the second half of fiscal 2027. Both products are in early access, and we're seeing tremendous interest from customers. These new solutions will provide organizations the ability to enforce AI policy, both at our exchange and the endpoint, enabling policy enforcement at the optimal location. We continue to innovate in this area at a rapid pace.

Next week, we are announcing the next major innovation on our platform with our new Agentic SecOps solution. Just as AI is increasing the threat surface, it is also stressing a human-driven traditional SOC approach, where remediation can take days or weeks. In contrast, our AI-first approach brings together our proprietary telemetry and Red Canary's decades-plus of experience in managed detection and response or MDR.

Our Agentic SecOps solution enables security teams to prioritize real threats and leverage specialized AI agents to detect, investigate and respond to threats at machine speed. We're driving closed-loop remediation in real time by integrating our Agentic SecOps with our Zero Trust Exchange, enabling customers to move with speed as the time between detection and exploitation has decreased from months to minutes. We will be launching our new Agentic SecOps solution with a webcast on September 9, which will be streamed on our website.

Our approach for securing users and nonusers and ensuring safe adoption of AI is resonating. This is increasingly evident in my conversations with customers and partners and is illustrated with a few customer examples. First, we had a notable 7-figure upsell Z-Flex win with a Fortune 500 transportation customer who deployed our security for AI portfolio to secure their full AI life cycle.

Our security for AI solution provides an integrated approach to secure AI use at scale. This includes discovery and management of all AI assets, including shadow AI and enforcement of safe access to approved apps. It also includes real-time prompt and response inspection to stop data leaks and threats like prompt injections and continuous red teaming assessments. This customer selected our security for AI solution over 2 major platform competitors and with this win, the customer's ARR grew to nearly $10 million.

In another 7 figures, Z-Flex upsell, a Fortune 500 semiconductor manufacturer expanded its adoption of the Zscaler platform to secure a company-wide rollout of Claude Cowork. After assessing several vendors, this client determined Zscaler security for AI was the only solution capable of securing the customer's AI adoption, including its endpoints, secure agent to agent communication and model usage in private and public environments. This is a great example of how our customers are expanding the use of Zero Trust Exchange to safely deploy AI agents and models as well as combat the threats created by Agentic AI. While we're in the early innings of security for AI, these deals give us tremendous confidence in our ability to expand this offering significantly over time.

Next, our leadership in data security is a powerful tailwind for our Security for AI business. As enterprises embrace GenAI and Agent AI, they're confronting a new wave of data exfiltration risks, including data abuse and over privileged access to data. Our in-line architecture, coupled with our Endpoint DLP and now Endpoint AI Security enhances our ability to enforce data loss prevention in the cloud as well as on the end point.

Securing data and AI go hand-in-hand as evidenced by the fact that 70% of our security for AI deals this quarter included our data security solution. In Q4, we also closed a 7-figure Z-Flex upsell win with a large global asset management firm. The customer selected Zscaler's data security posture management solution over a privately-held DSPM vendor. This customer chose Zscaler to address gaps related to sensor data discovery across its multi-cloud environment and data governance.

With this upsell, the annual spend of this customer increased by nearly 40%, reaching an ARR of $5 million plus. We're also seeing continued traction across our Zero Trust SASE solutions, including customers expanding their Zero Trust SASE deployments. For example, this quarter, we signed a 7-figure Z-Flex upsell with a Global 2000 financial services customer, who upgraded to Zscaler Private Access with AI-powered app segmentation for 120,000 users, increasing their ARR by nearly 50%.

We are also driving new logos. We closed a 7-figure new logo Z-Flex win with a Fortune 500 life sciences company that is deploying our Zero Trust SASE and security for AI platform across 75,000 users and displacing a legacy firewall-based SASE platform. This customer, led by a newly appointed CISO who is a 3x repeat Zscaler customer, chose us for our ability to deliver unified visibility and control across both enterprise security and GenAI. This is a great example of a new logo purchasing the entire Zscaler platform.

Customers are also increasingly starting the Zero Trust journey by securing nonuser environments. For example, we closed a 7-figure new logo win with a Global 2000 health care equipment manufacturer to deploy our Zero Trust Branch solution across the critical production sites to secure its OT environment, displacing an existing long-term legacy vendor. This is an example of a sizable opportunity Zscaler has within factories and warehouses to secure IoT/OT and provide Zero Trust device segmentation.

Zero Trust Branch simplifies customers' branch deployments by eliminating traditional branch firewalls, SD-WAN and MPLS networks, reducing operational complexity. It also minimizes the impact from an infected machine in a branch or a manufacturing plant by limiting lateral movement across the environment and therefore, containing the breach.

In another known user deal win, we closed a 7-figure upsell with a Fortune 500 aerospace customer. This customer was going through a divestiture and expanded its Zero Trust Cloud deployment between its on-prem and public cloud environments to securely migrate workloads. With this deal, the ARR of this customer grew by more than 30% to over $5 million. We're seeing tremendous momentum with Zero Trust Cloud. This quarter, we extended the solution by offering a managed service through Google Cloud, which can be configured in under 10 minutes, reducing the deployment time and operational costs significantly. This expands on our existing offering for AWS and enables multi-cloud flexibility.

The strength we are seeing in Zero Trust Branch and in Zero Trust Cloud is translating to meaningful momentum with Zero Trust Everywhere enterprises, those that have purchased Zero Trust Users, Zero Trust Branch and Zero Trust Cloud. We exited Q4 with more than 950 Zero Trust Everywhere enterprises versus over 700 in Q3 and over 350 at the end of fiscal 2025.

To summarize, AI represents one of the most significant opportunities in Zscaler's history. Our platform was built for this moment. Zscaler has always secured interactions between every user and applications, and now Zscaler secures interaction between every user, every agent and every AI model. We have a large and growing market. Adoption of our platform is expanding, and we are continuing to innovate across Zero Trust SASE, Agentic SecOps, data security and security for AI. We are well positioned to extend our leadership as the cybersecurity platform for the AI era, drive durable growth and create long-term shareholder value.

Now I will hand it over to Kevin to walk through the financials.

Thanks, Jay. We delivered strong Q4 results with revenue and ARR both growing 25% year-over-year, net new ARR growing 24% and non-GAAP operating margin reaching a record 24.3%. For full year fiscal '26, revenue also grew 25% and when combined with 23% free cash flow margin, our performance exceeded the Rule of 40, landing at approximately 49%.

Our growth engine continues to broaden beyond users with increasing contribution from non-seat-based solutions, continued Z-Flex momentum, record large deal activity and improved sales productivity.

ARR momentum remains strong in Q4. Excluding the contribution from our acquisition of Red Canary, net new ARR was $232 million, up 17% year-over-year, and total ARR was up 20%. Importantly, this marks a continued acceleration in net new ARR growth from 7% in fiscal '25 to 10% in the first half of fiscal '26 and to 17% in Q4. Red Canary exited Q4 with $141 million of ARR.

Total net new ARR was $246 million, up 24%, bringing total ARR to $3.8 billion, up 25% year-over-year. Performance was broad-based with strength across the Americas, EMEA and APJ.

We also continued to deepen enterprise adoption. In Q4, we closed with a record number of $1 million plus new ACV deals and the number of $10 million-plus ARR customers nearly doubled year-over-year. We ended the quarter with 785 customers generating over $1 million in ARR and 4,182 customers generating more than $100,000 in ARR, growing 18% and 20% year-over-year, respectively.

As our platform expands beyond users to protect branches, workloads, AI applications and AI agents, our monetization model is also expanding. In Q4 and for the full year, our non-seat-based metered usage solutions delivered approximately 30% of new and upsell ACV. ARR tied to these offerings grew more than 100% year-over-year.

Turning to revenue. Q4 revenue was $898 million, up 25% year-over-year and 6% sequentially, exceeding the high end of our guidance. Growth was broad-based across the geographies. The Americas accounted for 57% of revenue and grew approximately 30% year-over-year. EMEA accounted for 27% of revenue and grew approximately 17%, and APJ accounted for 16% of revenue and grew 23%. For full fiscal '26, revenue of $3.4 billion grew 25% year-over-year.

Excluding Red Canary, revenue of $3.2 billion grew 20% year-over-year. Red Canary contributed $144 million of revenue in fiscal '26. The Americas accounted for 57% of revenue and grew approximately 31%. EMEA accounted for 28% of revenue and grew approximately 16%, and APJ accounted for 15% of revenue and grew approximately 23%. Remaining performance obligation, or RPO, of approximately $7.4 billion grew approximately 27% with approximately 45% classified as current RPO.

Turning to go-to-market. We are pleased with the continued strong sales execution. In Q4, we delivered double-digit sales productivity growth and achieved our highest quarterly productivity ever and the highest annual productivity since 2022. This reflects continued improvement in our account-centric sales motion and our ability to drive broader platform adoption with customers.

Entering fiscal '27, our priorities are focused on deepening relationships with existing customers, accelerating platform adoption, improving new logo execution, expanding coverage in key segments and increasing partner-led contribution. These priorities are especially important as the rapid proliferation of frontier AI models increases customer engagement at senior levels of the organization as Jay discussed earlier.

To support new logo growth, we are expanding coverage through both direct and partner-led motions. For example, we recently expanded our partnership with Carahsoft to further penetrate the commercial and SMB segments in the U.S. through 100% channel-led motion. We are also adding dedicated new logo sales executives focused specifically on pursuing new enterprise customer opportunities.

Z-Flex remains an important part of our go-to-market strategy. It provides customers with multiyear commitments, the flexibility to activate or swap modules without starting a new procurement cycle while also providing premium deployment assistance and support. We saw strong momentum again this quarter with Z-Flex, driving meaningful upsell, shorter sales cycles and greater forward visibility. In Q4, Z-Flex generated over $770 million in TCV, up more than 60% quarter-over-quarter.

For fiscal '26, Z-Flex customers saw an ARR uplift averaging nearly 30%. For fiscal '26, Z-Flex generated more than $1.7 billion in TCV, underscoring customers' long-term commitment to Zscaler. Two recent examples illustrate the value of this model. In a 5-year 8-figure Z-Flex deal, a Global 2000 services customer increased its ARR by nearly 90%, crossing $5 million this quarter and driven by expansion with existing products and adoption of new products, including our security for AI and Zero Trust Cloud solutions.

In another example, an existing 7-figure ARR Global 2000 retail and wholesale customer increased its annual spend with us by 140% in a 3-year 8-figure Z-Flex deal. This customer expanded adoption across 9 existing modules and adopted 6 new modules, including our security for AI solution.

Turning to operating performance. We delivered strong profitability while continuing to invest in the business. Non-GAAP gross margin was 80.2% compared to 79.3% a year ago. The year-ago period included a onetime deployment of a large private cloud in a government customer's data center, which included a hardware component that carried a lower gross margin profile. We also delivered significant operating leverage in the quarter. Non-GAAP operating income was $218 million, up $60 million or 37% compared to $159 million last year. Non-GAAP operating margin was 24.3%, up 220 basis points year-over-year.

For the full year, non-GAAP gross margin was 80.3%, up 20 basis points year-over-year. Non-GAAP operating margin was 22.9%, up 120 basis points year-over-year. This performance reflects the strength of our business model, disciplined execution and our ability to deliver durable growth with expanding profitability while continuing to invest in the significant market opportunity ahead.

Turning to the balance sheet. We ended the quarter with approximately $3.5 billion in cash, cash equivalents and short-term investments and $1.7 billion of debt. In Q4, we generated $279 million in operating cash flow, and CapEx was $200 million or 22% of revenue. This brought our full year CapEx to $277 million or 8% of revenue and coupled with capitalized internal use software of $73 million resulted in free cash flow of $779 million for the full fiscal '26 or a free cash flow margin of 23%, down from 27% last year. The year-over-year declines reflect the timing of cash collections and CapEx expenditures.

Looking ahead, I want to provide an update on our expectations for capital expenditures to support our growth. As I shared last quarter, we are seeing higher prices and tighter availability for memory, storage and processors. These components support our data center infrastructure and our Zero Trust Branch appliances. In Q4, we opportunistically accelerated certain purchases where data center equipment was available. As a result, fiscal '26 CapEx was 8% of revenue, consistent with the expectations we provided last quarter for CapEx of high single digits as a percentage of revenue. We expect CapEx to remain elevated during fiscal '27 due to higher component pricing, especially memory. We'll continue to monitor our costs and share regular updates about the impact.

To provide additional capacity to support our AI and growth initiatives, we are strategically reallocating resources through a workforce restructuring. This action is expected to affect approximately 3% of employees and result in restructuring charges of approximately $30 million to $33 million.

Turning to guidance. Let me provide our outlook for Q1 and full year fiscal '27. As a reminder, these numbers are all on a non-GAAP basis. For the first quarter, we expect revenue of $935 million to $939 million, approximately 19% year-over-year growth; gross margin of approximately 80%; operating profit of $215 million to $217 million, up approximately 25% to 26% year-over-year, representing a 23% operating margin; net other income of approximately $33 million; and earnings per share of approximately $1.15 to $1.16 per share, assuming a 21% tax rate and 170 million fully diluted shares.

For the full year fiscal '27, we expect ARR of $4.396 billion to $4.426 billion or year-over-year growth of approximately 16.6% to 17.4%. For net new ARR seasonality, we expect approximately 37% of net new ARR in the first half of fiscal '27 with 15% in Q1 fiscal '27. Revenue of $3.908 billion to $3.938 billion, reflecting year-over-year growth of 16.6% to 17.5%. Gross margin of approximately 80%, operating profit of $924 million to $932 million, up approximately 21% year-over-year and equating to an operating margin of approximately 23.7%. Net other income, approximately $140 million to $142 million. Earnings per share of $4.86 to $4.90, assuming a 21% tax rate and approximately 173 million fully diluted shares and free cash flow margin of approximately 23% to 23.5%, reflecting CapEx, not including internal use software in the low teens as a percentage of revenue. We expect free cash flow margin to be seasonally stronger in Q1 and Q4, reflecting timing of CapEx and cash collections.

Looking ahead, we are excited by the opportunities we see to continue scaling our rapidly expanding AI Security portfolio, accelerating Zero Trust Everywhere adoption and growing our data security revenue.

In summary, we are pleased with the results we delivered in fiscal '26. We achieved 25% year-over-year ARR growth, record operating income and operating margin. We also saw continued momentum with Z-Flex and closed an all-time record number of $1 million-plus ARR deals. I'm excited about the substantial opportunity ahead in fiscal '27. We are confident in our ability to continue to drive profitable growth across multiple vectors, including product innovation, go-to-market and customer expansion and creating value for our shareholders.

To learn more about our strategy and plans for the future, please join us at our upcoming Investor Day on October 6 in New York City.

I want to thank our employees, customers and partners for their continued support. With that, operator, you may now open the call for questions. Thank you.

[Operator Instructions] Our first question comes from Saket Kalia with Barclays.

Okay. Great. A nice close to the year. I'll keep it to one, but maybe address it to both of you, Jay and Kevin. It's great to see the higher growth rate for next year in ARR. Maybe the question is, how are we thinking about the impact of some of the sales churn here that we talked about last quarter in terms of productivity? And do we feel like the sales churn here has stabilized as we go into '27?

Let me start. Over the past few quarters, we have gone through transformation of our sales organization where we expanded from opportunity-led sales to account-focused sales. The strong results of Q4 kind of show that the transformation, the changes we've driven are working well. If you look at the sales productivity, actually, the sales productivity in Q4 was the highest. I think we are very well positioned with a strong sales team with good channel partnership and a very strong product portfolio as we enter fiscal '27. Kevin?

Yes. Thanks, Jay and Saket. Maybe just to reemphasize the point. So Q4 marked the highest productivity quarter that we've ever seen and fiscal '26 was the highest annual sales productivity in 4 years. So I think we're really positioned and set up well going into fiscal '27. Just as a reminder, we did have 2 sales leaders depart last quarter. One was a geo leader and the other was a vertical leader. We backfilled and internally promoted the vertical leader. Since then, a new leader has accepted the offer for the geo position. Just keep in mind that certain geos may take longer to onboard than others.

And finally, as we think about the shape of the transition, we expect that the leadership transition will play out in the first half of this fiscal, and that is reflected in our guidance. And I am excited about how this leadership team is coming together as we think about augmenting our existing team into this year.

Our next question comes from Fatima Boolani with Citi.

I was hoping to have you unpack some of the discrete drivers of that organic net new acceleration in the quarter. Provided that this is your fourth fiscal quarter, it's an abundant period of renewals and expansionary behavior from customers. So I'd love to kind of how you stack rank between some of those inputs and how did you feel and think about your new logo activity in the quarter? Just wanted to get a sense of the most sensitive drivers of upside and strength there relative to what we were looking for.

I'll start, and Kevin can add on. Q4 was a very strong quarter from almost all measurements. You saw all the metrics. We did very well. Our product portfolio has become pretty broad. Zero Trust Everywhere has been driving a lot of our sales. AI security peers are growing out there. And one of the things AI security needs Zero Trust architecture, Zero Trust solutions. Our solution like Zero Trust Branch did extremely well. Data security drove a big part of our business. And even the young offerings of security for AI has grown very nicely. So all products actually did very well. Kevin?

Yes. Thanks. So maybe just a couple of other points to emphasize. We did have broad-based strength in the quarter. So I think that's number one to reflect. NRR, although it's not a metric that we provide each and every quarter, it was consistent each quarter this year at 115%. So we've seen very consistent performance in terms of our upsell motion that I think is important to appreciate. Maybe one other comment I'll make is, as Jay mentioned, the early success with our security in AI. Obviously, it had a strong performance in the quarter. But another optimistic point of view is that we also saw a 75% increase in pipeline in security for AI as we think about what that sets up for '27.

If I may add 2 more points, from a new logo point of view, our Fortune 500 penetration moved up from 45% to 50%. That's pretty remarkable. And also worth noting is the $1 million-plus deals. We had a record quarter for new million ACV teams.

And Fatima, my reference to the increase in the pipeline relative to AI was quarter-over-quarter sequential, not year-over-year.

Our next question comes from Roger Boyd with UBS.

Awesome. Jay, I wanted to touch on the competitive environment you're seeing with some of these AI security wins. And I want to maybe pick on the 7-figure upsell you called out with an airline company where you won over 2 other platform competitors. Can you just talk about what you're seeing in these bake-offs? Presumably, companies are taking these decisions pretty seriously and trying to make strategic bets on which layer of security they see as best positioned to secure AI. But I'd be curious to see what you're seeing and what's causing you to win these large deals.

Yes. AI security is in demand, and there are probably lots and lots of AI security companies, probably tons of start-ups as well. So one of the things we see out there is our customers tell us that they don't want 1 more or 3 more AI security products. They're looking for an integrated solution, not only for AI security, but also that works with the rest of security as well. So this transportation customers that we highlighted, they are a Zscaler customer for Zero Trust Everywhere.

And it's natural for them to say, if I need to expand into AI security, then tomorrow, I need to be ready for my agentic communication, where Zero Trust Exchange to be able to ensure policy for agents. We were the natural choice for that to happen. And yes, every company is offering AI security, but they all come from different sides of it. Our differentiation is we have an [ interior ] solution for asset management to communication to actually relationship figuring out the access graph as well as agentic communication actually. That's really what's setting us apart.

Our next question comes from Richard Poland with Wells Fargo.

I just wanted to get a sense for unpacking some of the verticals in the quarter. I guess when we think about federal, I know last year, federal was a little bit softer than we would have hoped for. But just as we think about this quarter and heading into fiscal year-end for federal, anything to keep in mind there?

Yes. I'll take that quickly. Look, the federal business contributed high single digits to new ACV in '26. That was very similar to '25. Expectations for fiscal '27 are similar. There's nothing in particular that I would point out. Obviously, earlier in the year with [indiscernible], it was a difficult environment, but we've seen it perform fairly consistently with the last several years.

Our next question comes from Joseph Gallo with Jefferies.

Nice job on a quarter. I was just hoping if you could break out the Symmetry contribution to net new ARR in 4Q. And then just as a part of that really strong strength in non-seats in fiscal '26, what are the expectations as we go into fiscal '27? Because you're obviously offering more and more there. So is it fair to expect a larger mix of the new business comes from that non-seat?

So Symmetry is a very innovative technology that essentially connects the dot between different entries accessing different data sources as we have seen proliferation of AI agents. This innovative technology becomes extremely important. This was essentially acquired as a technology company with little -- very little revenue for it. But it's becoming an important part of Zero Trust Exchange, not only for agents, but Zero Trust for cloud workloads and Zero Trust for devices because understanding relationships of entities to what they access is foundational piece. That's where it's fitting in.

Regarding non-seat at a broader level, the comment I'll make is while early on, our products like ZIA for users, ZPA for users were largely user-centric. A lot of our offerings subsequently have not been user-centric. For example, Zero Trust for cloud workloads, it's all about workloads. It's essentially consumption based. Zero Trust Branch, which is largely about branch devices, IoT/OT devices communicating is all about non-seats. Data security, some of the modules are [ seat-based, ] others are data volume-based. And almost all of our AI security products or security for AI products are consumption-based linked to queries, essentially transferred to token consumption fundamentally. Kevin?

Yes. So with respect to Symmetry, again, to Jay's point, it really was a technology and talent acquisition and provided the access graph technology for us. The results are immaterial to our results, and nothing else to call out there.

Our next question comes from John DiFucci with Guggenheim Securities.

I have -- I apologize. I have a 3-part question, but I promise, it's all related. You said you're adding new enterprise reps to go after enterprise accounts while also restructuring, and that's going to affect 3% of your employees. Can you give more color around this? One, what kind of employees will be affected? And what will be the net employee count effect? And then also, where are these new reps needed, is it specialized sales or will it be broadly across your geos and product? And finally, what's driving this, is it an inflection point in demand that Jay talks about AI or is it just as you said -- you also said here, you're seeing very good sales productivity, so now it's just the time to be hiring more to keep on growing?

Yes. Let me start. So first of all, the reduction you pointed out, it's essentially rebalancing a reallocation of some of our resources with better leverage and also better investments in the AI area we are looking at. That's fairly straightforward. And if you look at adding our sales resources, when a company is growing at a pretty good pace, it's natural to add resources. We have essentially almost always done it. Now the question is where do we add the most of the resources.

We do have specialty teams and we have account execs. Our specialty teams have been making a pretty meaningful contribution. I'm extremely proud of it. Our CRO, Mike Rich and his team evolved the idea of take-up teams that we have started about 3 years ago. It has become a very good specialty team, and we are adding resources to specialty teams. But also in geos too, a number of accounts execs are being added. One of the areas for addition of account execs is probably the enterprise. If you look at the top end of the enterprise, we're pretty well covered. As you come down in the enterprise market, our coverage has been thinner. We are making that coverage. We are also investing on some of the channel resources because channel is helping us expand in the mid-market and like. Kevin?

Yes, John, the only thing I would add is the restructuring is about 3%. So it's modest, about $30 million to $33 million in restructuring charges. Maybe just to also double down on what Jay mentioned, Mike is adding people to the organization, both to address opportunities we have with new logos as well as upsell. We have about 4,600 of 20,000 targeted companies. So there's a big opportunity just with new logos alone. And so Mike does see an opportunity to do that through dedicated new logo hunters along with reps that hold both existing customers and prospecting territories. So there's opportunity for us to continue to expand across both of those dimensions.

So just to be clear, guys, and thank you for that detail. It's more about, hey, listen, our sales force is getting really good and productivity has gone up. So in order to continue to grow, and you're bigger, too, the law of large numbers, you just need to hire more salespeople. But it's more of that than it is when Jay talked about AI changing demand for security out there. Is that -- am I reading you right?

I don't think it's either or to be honest. I mean we certainly are seeing strong tailwinds with respect to AI broadly, whether it's specific to security for AI or just broad-based interest in Zero Trust and our existing kind of core Zero Trust Exchange. So I wouldn't limit the discussion to simply just generally continuing to hire to grow. I think there is a unique momentum at the moment with respect to kind of this AI tailwind.

Our next question comes from Brian Essex with JPMorgan.

I guess I want to take the other side of the coin relative to what Fatima asked. And I want to ask about the health of the Red Canary business. I know when you entered the year, there was some question of what the renewal experience would be on that platform, and you've now had a year worth of renewal experience. I think it's performed relatively well. But you also commented I think at the beginning of the year that some of those customers aren't your typical Zscaler customers.

So now that you've kind of like had a year of experience, what is your expectation for the growth of that business? And I totally understand the rationale behind the IP acquisition and the value that's going to provide to the products being released very shortly. But I just want to kind of get a baseline of what is your view of the health of that customer base, the renewals and growth of that base business going forward, so we can kind of gauge what -- how to interpret the outlook for fiscal '27.

I will start with a broader comment. The acquisition of Red Canary was, number one, to make sure we have agentic technology that could become part of our SecOps platform. Number two, we didn't have any SecOps expertise. Red Canary had 10 years of experience about understanding how our SecOp runs, and that was important for us. In that area, we've done a pretty good job of integrating their technology with our technology, and that's becoming an important part of our SecOps solution. Regarding financials, Kevin?

So look, fiscal '26 was really focused on integrating the Red Canary technology into our Agentic SOC solution that's launching next week. When we closed the acquisition a year ago, we did not know what portion of their business was going to be durable. So we took a conservative approach in how we picked up ARR. Quite frankly, we're pleased with the ability to maintain the book of business while going through the integration. As I mentioned last quarter, as we think about fiscal '27, we are integrating their technology. It is a combined integrated offering that launches next week. And so as it relates to Red Canary, we're not expecting any net new ARR contribution. All of that will show up in the integrated solution going forward.

Very helpful. Do you think the churn is pretty much in the rearview mirror at this point in terms of any churn on that platform that may have happened?

I mean, look, we've talked, I think, at nauseam about the different churn rates that MDR businesses experience discretely different from ours. And that was -- that played out that way. We did see elevated churn in Red Canary's business in '26. Despite that, we were able to continue to maintain its book of business, as I mentioned. I would expect that as it relates to accounts that are up for renewal that we'll continue to see the higher, more traditional churn rates of an MDR business, not our rates, but that has all been contemplated in the guide.

Our next question comes from Meta Marshall with Morgan Stanley.

Great. I wanted to ask a question about the Z-Flex traction that you're seeing. And just as you continue to expand the platform with AI security and Agentic SOC, do you envision kind of mandating Flex plans more with some of these new products? And just how is it impacting sales cycles relative to the traditional sales approach?

I'll start. So Z-Flex is not about mandating. Z-Flex is about providing flexibility. As our platform is getting bigger, many times, our customers will look at evaluating product A or B or C or D, and they're not sure which one do they want, and this will take longer time. By giving them the flexibility there, you can start with a certain number of products, you can swap other products without going through, again, typical procurement level. This is one of the biggest things they needed. And sometimes, they wanted the ability to ramp because if they bought 6 products rather than 3 of them, they want some ability to ramp that became part of the Flex deal. And they also want a longer duration.

The customer engagement with us are not transactional. They're generally long term. Once they deploy us, they invest. They want to stay with us for the longer time. And they also want the ability to buy additional product from a rate card. All those things are very good for business, for customers and good for us as well. But it's true as our platform becomes bigger, all the products will be available as a part of our Z-Flex offering.

Yes. And one of the other advantages, it was in my prepared remarks, is it does -- it eliminates the need to go through new procurement cycles every time a customer wants to adopt and implement new modules or features on the platform. So you go through that discussion once. They have to complete flexibility the choose what's appropriate for them to use with their business at different points in time. And these are longer-term commitments, so we can offer that level of flexibility. So we've been very pleased with the momentum with Z-Flex. And as I mentioned, we ended fiscal '26 with more than $1.7 billion in Z-Flex bookings. So it's been very, very well received.

Our next question comes from Shrenik Kothari with Baird.

Our next question comes from Ittai Kidron with Oppenheimer & Company.

A solid finish for the year, guys. Kevin, I want to dig into your outlook for fiscal '27, specifically on the ARR front. I would love if you can give us a little bit more insight into the puts and takes that you've taken into account into that ARR guide. I mean clearly, you've had very good momentum here with net new ARR. If my math is right, that declines to 4% at the midpoint for your guide for fiscal '27. So would love to kind of get a little bit more kind of color as to what's included with respect to the go-to-market, the new products and any other components that you think it's important to call out. Appreciate it.

Yes. Thanks, Ittai. Maybe at the highest level, what I would say is we're very optimistic with the momentum that we do carry into '27, especially after a 17% net new ARR growth in Q4, excluding Red Canary. I am considering the time it will take for the sales transition that we've talked about, both in terms of the geo and the verticals as well as the pace of the uptake of the integrated SecOps solution.

In addition to that, look, as I mentioned, we have had consistent NRR. I mentioned each of the quarters of fiscal '26, it was $115 million. So I do expect that to continue. We also have an opportunity to accelerate new logo growth, which we've talked about. We're adding dedicated new logo sales executives, specifically focused on enterprise customers. We also have, from a product point of view, a larger opportunity to land new logos with a variety of products, very different than was the case years ago. And so I'm pretty optimistic in terms of the different dimensions that we think about for '27 and opportunities for us to continue to grow.

And AI security is adding further tailwinds.

Is the AI security contemplated materially into your outlook because a lot of that portfolio still needs to come?

Yes. We haven't specifically called out the contributions from security for AI in the guidance. But as I mentioned, AI is very quickly becoming a strong and durable tailwind for the business. And it is driving demand, not just for security for AI solutions, but Zero Trust Everywhere and data security. And we expect that these drivers will persist in '27.

Our next question comes from Gregg Moskowitz with Mizuho.

Jay, as you mentioned, you're hosting a launch event for Agentic SecOps next Wednesday. Quite frankly, a lot of time has passed since you acquired Red Canary and now that the day is upon us or almost upon us, it would be helpful to hear just a bit more from you as to what might be unique about your Agentic SecOps and what it will unlock for your customers.

Yes. Very good question. So our customers have been talking to us for the last few years. They're telling us that we have the best data, best telemetry across the network because sitting in line from endpoint to sitting on the endpoint and cloud as we're sitting on cloud workloads as well. This is resulting in over 750 billion transaction logs per day. That's a starting point of really good SecOps. So our customers are saying, we have to send this data to another vendor. You've got the data right there, why can't you give us more meaningful information that's needed for SecOps, which is natural. We should be doing that. That was number one driver.

Number two was, we observed over the years that there was a first generation of SecOps solution. Then second generation came where they're still human-centric but automation-driven. They're essentially human-driven. We saw the opportunity to really build a SecOps solution truly agent native, where agents are driving, detecting and able to investigate this thing at machine speed versus a traditional solution. I think it's an opportunity to disrupt traditional SecOps solution, even the one that they call themselves next-gen SecOps.

And the last factor was closed-loop remediation. These days, the time between a vulnerability being discovered and exploitation and shrinking typically take days or weeks for a typical SecOps solution to get all that telemetry, do all the detection and planning and really take an action. With Zscaler, since we have most of the data, we could do it within minutes. Those were the big drivers for us. We have a large number of customers who are waiting for our solution to really be deployed. So we are pretty excited about it. I think we'll have a meaningful opportunity for us. In the first half of the year, it's going to take some time to take off, but it should start contributing in the second half and then fiscal '28.

Thank you. I would now like to turn the call back over to Jay Chaudhry for any closing remarks.

Thank you all for joining us today. We hope to see you at one of our investor conferences. Thank you again.

Thank you. This concludes the conference. Thank you for your participation. You may now disconnect.

Watkins, SVP, Investor Relations and Strategic Finance. Hello, everyone. Welcome to those of you joining us here in Las Vegas at Zenith Live and to those of you who are joining us online. We really appreciate you making the time to be with us here today. Before we begin, I'd like to remind everyone that today's presentation contains forward-looking statements within the meaning of the safe harbor provisions at the federal securities laws, including statements regarding our future financial performance, business strategy and market opportunities.

These statements are subject to risks and uncertainties that could cause actual results to differ materially from those projected, and we take no obligation to update them. For a more complete discussion of the factors that could affect our results, please refer to the risk factors described in our most recent filings with the SEC, including our annual report on Form 10-K and quarterly reports on Form 10-Q.

Okay. With that out of the way, we have a great lineup for you today. So in a minute, Jay is going to kick us off and take us through the Zscaler platform. and also go through some of the new solutions that we announced here today, including those that are securing AI. I know that is an interest for many of you. So we'll be sure to hit on those. Then Dave will come up and host a panel with 3 of our customers that we're really fortunate to have with us here today.

They're going to take us through some of their security challenges their journey with Zscaler and also take time to answer your questions. So start to think about what you might want to ask. And last, we'll finish up with plenty of time for an executive Q&A. So you can make sure to get all your questions answered. Okay. We're ready to get going. With that, it's my great pleasure to introduce our Founder and CEO, Jay Chaudhry, Jay?

Thank you. All right. Good afternoon. Good afternoon. Great. So as Kim said, I'll give you a high-level view of our platforms, on the offerings, what sets us apart from others. -- as those of you who attended the morning keynote, they may be a little bit duplicate as we are broadcasting session.

I want to make sure that remote attendees also have a big picture view of it. So overall, if you think about the opportunity that gets me excited is the massive market opportunity. It has been growing over time, and I'll walk you through how we have over $120 billion serviceable addressable market for us. And the need for cyber, the need for the solution we offer and talking about the architecture, what sets us apart and then a big picture view of the overall platform and close with some of the financial strengths.

Let's jump into the TAM. I think when we had Investor Day, we'll do bottom up, do some more analysis of it. This is kind of built upon the market size than we had shared with you before. On 0 trust every year, which is not just the users, which is the cloud and branches, you look at all those things together. It's about $65 billion SAM, pretty sizable.

We lead this area significantly, especially in the user side of it. The cloud is a great opportunity to disrupt the traditional virtual firewalls in the cloud and branch is an exciting opportunity to eliminate traditional wide network, traditional way of doing security inside the plants and factories. Data security is an ever-growing market as more and more data gets created and more and more on data sets out there.

And with AI, the data loss becomes a bigger challenge. We see this as an ever-growing opportunity for us. Agentic Ops, essentially is largely around SecOps and a couple of other areas like IT operations. But this market is sizable in the early stages market, but we have a chance to disrupt it. And securing AI is a brand-new market segment. And we got some serious momentum. We set up AI security as a start-up with a Zscaler to really build these products. And I couldn't be happier with the pace which we are building and developing these products and the traction or the interest we're drawing from our customers.

Let's look at -- the need -- you read all this stuff out there every day. So I don't need to walk you through all the stuff. But every day, every week, there's an issue that's happening. Somebody tried to embrace AI copilot loses data, open Clark, or credentials out there or some agent delivered some e-mails or they deleted some production database.

And a lot of these things aren't even has they are actually lack of policy, lack of controls, lack of guards. So you combine the cyber part of it with some of the guardrails net we built and wanted to make sure AI can be used reliably and effectively. It's a huge, huge need. As I talked to so many CIOs and so many CSOs. The #1 message comes from them is -- we have identified a few pilot programs.

We are ready to roll out. We have built some agents. I'm uncomfortable because the governance and controls aren't there. This is an interesting challenge everyone is facing and distribute from the exchange solutions will come in because think of the falling way. Where were some of these guards and controls and oil-based access, literally as a part of the application. You as a user, went to the application, application control what you could do and then application talk the data.

And now you can bypass the whole application, you go directly to the data. Where is governance, where is control? Where is all this stuff? And interesting challenge. This is where us being in the middle of it to really do policy governance that type of stuff will become extremely important.

A number of you may have seen this white paper that Anthropic published about a week ago, a Zero Trust or AI agents. As a read, I was wondering, did my Mark and team write it. It literally felt like what we advocate what we believe in the story was very simple for agents to work successfully. You can't let them roam and on the network. I had done network security. I have a firewall here and I have a firewall it doesn't really work. You really need to treat every agent is untrusted entity.

And through some policy controls, you need to make sure they only talk to right areas. And so that's important. You've also been reading about methods. I mean, so much has been talked about Metros is unbelievable. We have been part of the Glass wing program from day 1, early March time frame. We have been using it. It's pretty effective. It can find a lot of vulnerabilities.

The interesting challenges are being -- how do you fix them okay? Enterprises already have a large number of unmitigated unremediated vulnerabilities. Metros or for that matter, OpenAI is GPD 5.5 or purpose 4.7 or 4.0. They're all pretty sophisticated. They're going to give you 5x more -- what do you do about it? The answer is not that you're going to double, triple down on just patching. You will never get out of doing patching itself.

So the answer is if opportunities get discovered, they're not passed. It's natural that there will be more breaches than we see today. Then the next level question that CIOs would ask is, how do I -- what else can I do to minimize breaches? I know patching goes only so far? And number two, if we got breached, how do we minimize the impact of those breaches.

And that's where we actually fit extremely, extremely well. Number one thing our customers are doing to prevent breaches is hiding their applications, eliminating their attack surface. In the firewall wall, you're out there, you firewall. You can check VP and all the stuff out there, you scan and you see all of these things out there.

The way Zscaler was built. Your proxy or which you're hidden behind us. Nobody knows where you are. Number one 'thing we can do. And our customers are busy working with us doing that. And that's also leading to some of the upsell opportunities for ZPA and some of the deception technologies because they want every user to be able to do Zero trust when they access any application.

The second part ends up being stopping natural moment. Otherwise, a single infector machine in one branch can infect everything else out there, not a good idea. What if that could be contained in the branch itself -- we do that extremely well. So those are the 2 best defenses that our customers want.

And we are working with the leading model companies. As I mentioned, we are part of Anthropic Project GlassGreen. We're also part of OpenAI, they break. And it's good to work with them because they actually are helping to actually bring the applications to the market, we become an important partner to make sure those applications can be securely used.

So the whole notion that these model companies are going to eliminate our SaaS surplus will happen or cyber goods disappear. If you dig into Methos a little bit, so Methos finding more vulnerability, that means there's more need for providers like CSC. So the notion that these guys will go and do that stuff is really unfounded. The other part is a provider like Zscaler, we have a global infrastructure around the globe.

There's 160 on exchanges out there, there's a public, and there are quite a few private exchanges major customers, okay? An agent is not going to go and create all of the infrastructure for your connection, network traffic routing -- all that stuff is a fairly complex and sophisticated area. That's why we feel like our need for us will grow because the more agents you have, the more policy enforcement, more in-line inspection you'll need, which is important because then we can help our customers and it creates a revenue opportunity for us.

So the platform is meant for. This seems like Zscaler moment. I mean we built this platform for stuff like this. When we started the company, we built, we evangelize this stuff. When COVID came then the market realized said, "Oh, we need something like Secure. That was a big moment. We think this moment is almost like COVID because, in fact, it's even bigger from a cyber point of view as everything is online, everything is digital. So that's a platform we built.

Just to refresh your memory on what we've built, what we're doing. On the left side is what you see. This is a typical corporate network. Everything connects to everything. Every office connects to every office, every IoT device OP device is connected because otherwise, you can't communicate and when you do VPN sitting at home, you're all part of the same network.

Your network extends to every household are -- and this is primarily the biggest reason of the problems -- and all these firewalls sitting out there, they become fairly porous. They try to do segmentation with it. And when they find oh, this source IP to this destruction IP? Well, these 3 users need this. They need this.

Then you know what the rule becomes any to any. It essentially becomes an open thing. That's why we need to move away from the world of firewalls to the Zero Trust world, where literally everything is literally in Island. They simply connect to Internet. We are the exchange with the switchboard, making sure the right party can talk to right party only, that's fundamentally what we're doing.

So when people talk about this vendor, that's SC vendor, all the SC vendors doing spinning up virtual firewalls in the cloud fundamentally. There's no Zero Trust in it, okay? If people think that they don't need Zero Trust and the firewalls are fine. And part of the reason why firewall companies will not do real Zero Trust is because it cannibalizes all the firewalls.

When we go in, tons of virals are taken out. It's not in the best interest. It's just like telcos were fighting, not eliminating MPLS. They'll go out and tell the customers say, "Don't do this because there is no quality of service. None of that is there. Guess what secular forces are very thoughtful. Similarly, I believe that the Zero Trust is a secular trend.

That's the only approach that's needed. That's what we pioneered and that's where we have for meaningful lead and others can't even try to do it because it's not in the best interest. Okay. So here -- so the other thing, I often get asked the question and say, "Oh, say this, SaaS this, I see this. When others talk about SaaS, they talk about secure access to users.

Now area we pioneered when we started with Zero Trust, any user can have access to any application or anywhere without being on the network. We are understanding that we moved on to do Zero Trust branch. Every branch is in Island, very important area, then doing 0 to inside the branch for every device. So an infected IoT device in the plant or in a factory can, in fact, other devices is very important, otherwise to Imagine if a plant goes down, it's an important area.

And Zerocloud is about cloud workloads, fascinating story amount to workloads in the cloud will keep on going. And AI will further accelerate the development of these workloads out to. How is this cyber done? -- east-west firewalls, no of we'll do a virtual firewalls. This source IP address can talk to this Destin IP address. Not very exciting, not very manageable and this when we come in and say goodbye to all these virtual firewalls, and we can do true Zero Trust in the cloud. Very exciting area and growing very well for us.

The most exciting announcement for us this week Zero just for AI agents. This is fantastic. Now as I said during my keynote, the rally about probably 70% of the pieces we need to 0 trust for agents who are already there. Think of it, agents or like people, the digital workers -- we already have technology to do that.

Agents are like code. We've done it for workloads. We've got all the pieces, the policy engine, the logging reporting and all, it's all there. I'll come back to cover that a little bit more. So all this Zero Trust everywhere is done to really achieve 4 key areas. Security of AI, how do we secure all the AI application infrastructure, data security, cyber protection and Agentic SecOps.

Let me dig a little bit deeper into each of these. Security of AI. This is what every customer is 1 looking for to start with. Every customer wants to know what AI assets do I have? Where are they? Do you have the end point? I mean using externally, for example, public AI applications, all the my private AI models, prime bedrock, whatever the case may be, all what's on the endpoint. We brought together all of this as one dashboard being able to give you a full view of all assets for AI, no matter where they are, okay? -- and along with the risks they post. It's important. Every company talks for having AI Asset Management. So an EDR vendor when they talk about it, they're going to tell what's on the endpoint, because that's what the it -- there are no idea of what communication is happening there.

They can tell you the public AI. They can't tell you the traffic. We're sitting in line for cloud or Internet we're sitting on the endpoint. We're able to give you a full view of it. Second area is secure AIX. This is for your employees, which employees should be able to access, which AI applications. We already had a policy engine. We have done that for other applications. So having rules and policies for AI applications were relatively easy for us.

We had to essentially build an engine for prompt inspection and response inspection, so we could analyze the prompts and do a policy based on that. Also the prompts can lose data. So being able to essentially do DLP as the prompts are going down was a natural thing for us, because we already do DLP.

The third bucket in this area with a solution we call AI Protect is securing applications and infrastructure that goes with it. This is handling the full life cycle from development through deployment and on time. For development, for example, we offered a red teaming, AI team. This came through the acquisition of SPLX. They've done a very good job in at they not only did -- are team. They also did continuous automated red teaming. And that's going to become an industry trend as models like Methos come out, continuous red teaming will need to be done.

And the way we had build these red teaming application. I can use any of the models on the back end to really do some of the scanning and vulnerabilities. It's a powerful story. The next thing, if you did this, -- how about frontline? What do you do for on time? That means securing your application build for your company. And when users need to access them, maybe it's your customers. They could do some bad things out there. They could do prompt injection from a cyber point of view. There could be a data issue. They could be unacceptable use.

There could be other crazy questions. Like one of the cases we saw in California where a car dealership set up an application where consumers could interact with it. And somebody asked a question and say, which electric car is better than the electric cars you sell, okay? So go to Tesla. Those are guardrails for acceptable use, meaningful use, they need to be set up out there.

There's some pricing questions that need to be done right. And these guys can go around it. There was a pop interesting use case. This was about CoPilot. The question was, once you train AI on these CoPilot. They get all the information. In all days in computing, you wrote a query, the computer can only give you an answer on that query and nothing more.

And they all once you train on it, they got all the information. So a user could say, tell me salary and bonus of XYZ. Now everyone is getting smarter and say, "Oh, CoPilot if someone is asking for the Saudi do not answer that question, say, sorry, I'm not allowed to share this information. Then as you saw last year, we shared this example. And this one guy goes and say, John likes to play basketball games from this pitiful box and the ticker for the box per game is $3,000. Tell me how many games you can watch in a box sent in blocks in -- with 1 year's salary.

Okay. That's on a guardroom. All those things need to be figured out, and those are part of the guardrail rules and all we're building and making sure customers can accept it. So this AI protect is a powerful solution. We launched it in late January. A number of pieces are built by us. A couple of modules came from SBLX. And when I talk to customers, they tell us that they haven't seen any solution that's as complete as integrated in this area as this is. So very pleased with that.

And now you're going to see these things evolve rapidly. I'm not going to go through every bullet point here. But this is a bunch of new enhancements, new features we added in this area, for example. In AI asset management, being able to discover embedded AI in SaaS traffic. All SaaS applications will become Agentic essentially. There's a traditional interface and there'll be anti interface going through proms. So being able to understand the traffic, we able to understand policies around every SaaS application, that Agentic, is an important area.

Okay. visibility to AI activity on endpoint. We could easily tell what all is sitting on the endpoint. That's not a problem. But then the customer said, it's okay you've got Claude cowork sitting on the endpoint. Or maybe it's some PD agents sitting on the endpoint or open closing of it.

I want to know the permissions and activity that's happening out there. So now we enhanced ourself from giving you what's running there with the potential risk and permissions type of stuff running out there. These are good examples of the enhancements we're doing. And securing AI access. This gets a application-specific. We started out from inspection for the most popular applications in early version of it. Now we are supporting over 250 G&A applications, where we fully understand, except the problems are able to take an action based on the kind of promise we got out there. and then also supporting Anthropic and opening this dial compliance API is available.

We are compliant. We work with those APIs. In the third area, secured AI apps and infrastructure, a number of enhancements got done. -- stand-alone prompt hardening feature has got added to it. W are at teaming for MCP servers as MCB servers are being pronounced. You're going to keep on seeing the velocity of innovation, the last year development for us to make sure we stay ahead of anyone else in this area.

And then the next big thing is really Zero Trust AI agent. And this is 1 of the hardest problem to solve. This probably has bigger barriers to entry for any new entrants than any other area out there. Without going detail into it, it's essentially a version of essentially the Zero Trust exchange we built. -- but new things we needed was AI brokers, brokers for MCP, protocols, broker A2A had to be done. And then understanding the test has assigned, understand the intent, the risk being extract prompts, analyze it to intent and risk gets from their ability to do those things become important.

And then essentially a lower policy. It is the only real way to be able to handle it. Our view is that as agents get deployed, I mean, there'll be so much stays. It's not even the agent level. It's an invocation level that needs to be figured out. That means scale, that means granularity needs to be handled. We already do about 750 billion in transactions day. We think in this new world that we'll have to add a couple of zeros to it in terms of how much volume needs to be handled.

We feel pretty good about it, having the architecture, having the scale, have the experience to be able to handle in line is not a trivial thing. Anything you do in line, it better work, because otherwise, people can do the jaw. Anything you're reporting if it doesn't work, you don't get the right answer, people don't even know most of the time.

So that's why being able to have great response time, great scale, is fundamentally important to us, and this is where we are very, very well positioned far better than anybody else out there. And then the next problem, this is a fascinating problem. This is An example of a solution. Our customers want clamoring for on day, okay? We like to do that. We like to think what will be needed in a year or 18 months. I want to work on it today. I want to cook it, I want to refine. I want to get better than anybody else. So what is this? This is Access Graph. Now Access Graph is not just needed for AI. It's needed for other entities, too.

Smite Systems, a company we recently acquired solved this problem. It's a very hard problem. Many other things, if you ask me, asset management at all, is that a rocket science to do? Not really. You give a couple of quarters and 3 quarters, you're going to build it, solving the problem of taking all this metadata from all the application to understand in your enterprise, in your corporate network, which identity, which entity is reaching, which data source, which MCB server, which application, is a nightmare, because they just get on the network here.

You literally have information sitting in each application about what access happen. So Smite pulled all that matter data. Hold it out, this is billions and billions of data points. And now that the magic of AI to figure out these entities are accessing this, the data source, the data lineage, they call the graph because this is important. Now why is it important? Number 1 reason the customers but looking at symmetry and the large customers are looking at symmetry systems is to understand the lineage and from there, they need to understand data governance.

Even the issue of SOC compliance, how do you do SAW compliance? You have to prove that you've got all these controls in place. And those controls are actually through applications. when applications get moved aside, you go directly to data. How do you prove SAW compliance? They'll be impossible.

This kind of solution can help you prove what's talking to, what's going on. That's how Symmetry was positioned to sell it. When we -- as we saw this technology, then we said, well, this is great. But in the Agentic world, the data will go 10x, 100x. It will be impossible to do something without something like this. So first, we understand that graph. And then we use this information to apply policies for 0 for exchange to be able to see -- with this group of agents and have this group of applications or this group of data sources, that's what's exciting about it.

It just also shows the DNA of the scale is to be innovative, to do things far ahead of others and set the peso do. So this is our overall platform story. Then I won't go in detail out there, but the list of innovations in SAS is long. Every year, we do probably about 200-plus features, Hinda, ZIA, ZPA space out there. Browser extension, enterprise browser availability.

It's a specific use case, we needed some of the extension area. We did a tuck-in acquisition of SquareX, it did very well. B2B exchange is an exciting area. Supply chain is a big risk. The only competition we have in that space is 30-year-old side-to-site VPN connection. That's a problem, right? -- power segmentation of application. This group of users can do this group of applications.

We further made it simpler. We have been doing for some time. And Z agent framework. We created an overall framework. So where our agents can be for each product, each area, they're working from the same framework. I think we're going to cover some of that tomorrow and Adam session or agent for ZDX, for example, which can do all the stuff that people are trying to do, it's all automated. So you're going to see all products of Zscaler having the agent interface to be able to engage with our products across the board. Agentic SecOps new exciting area for us. We have been building the technology internally.

We got Red Canary technology. And as a result of that, we are going to really announce we had 2 product areas threat management, this is traditional security operations and exposure management brings together all the exposure area, your attack surface, your asset risk management and so on and so forth. And this is built on some pretty solid technologies and where we can take data from all kind of sources, including Zscaler sources -- and we've got a bunch of techniques and behavior-based analysis and all the mapping are done, context graph we create here to identify real threats. And this is just an exciting area.

This is also being launched this week, and you're going to see it grow every month as we move through the fast pace. Wrapping up the last couple of points. A number of questions have been asked for pricing user base versus nonuser-based pricing -- we started out very early on with seat-based pricing. But as we evolved, things have grown, for example, Zero Trust branch. It's based on a number of devices in the traffic as flows devices Zero Trust cloud workloads, a number of workloads and the traffic from those things. Data security at modules, only some of them are linked to a number of people. Others are based on the amount of data we are scanning, the data we are classifying -- and as you'll see, Agentic exchange, all of these will be based on agents, amount of traffic, which essentially leads to the token consumption essentially a consumption-based model.

So we're seeing our new business ACV coming from nonseat, nicely growing over time. About a quarter ago, we disclosed about 25% of the new ACV came from non-seat and last quarter in Q3, that number moved up to 30%. So we don't think we have any meaningful exposure based on seats because our model is expanding. Our platform is growing pretty rapidly.

And lastly, our scale, you know the numbers were to summarize, just to let you know, we crossed $3.5 billion in ARR. And we got plenty, on 20,000 enterprises we target about 4,500 are customers. That means remaining a prospect for us to pursue. This good opportunity for new logo. And also in the big areas AI protect that we just launched in January, right? We crossed $100 million over the last 12 months.

There are a couple of modules that were there like GenAI security is part of it, but a lot of stuff new. Most of the stuff is picking up -- picking up very nicely. Data security is growing very well, is going to keep on growing very well. We have $0.5 billion crossed in ARR and over 30% year-over-year growth. And Zero Trust every year is what sets us apart from others. Will set us apart for a long, long time. We started sharing with you the number of customers in Zero Trust everywhere.

About a quarter ago, that was 550, and now we crossed over 500 enterprises who do Zero Trust everywhere. That means they've got Zero Trust users, Zero branch and zero-trust cloud, okay? With that, we're going to start the next session. And this is our customer panel that the Dhawal Sharma is going to moderate.

Great. Thank you. Great second. All right. We will take about 20 minutes a discussion between us, the speakers that are here with me, our customers, and we'll then open it up for you to ask questions as well. Since we have 3 various teamed customers who have joined us here. Why don't we start with you, Wan and go around and get an introduction about you, your roles and how long you've been using Zscaler and what problems we are solving for you.

Wonderful. Thank you. Good morning, everyone. Thank you for having me. So Wayne Fudersky with Edward Jones, have been there 25 years. Deputy so responsible for enabling the firm securely responsible for all the enterprise security architecture products and solutions. Then working with Zscaler now since 2010. So really grown up with Zscaler. We use a lot of their key core products, EIA, ZPA, ZDX, CASB, browser and even touching now into the AI products. .

Jason Coler, been with Eaton for 10 years. I am deputy Ciso there. I've been a Zscaler customer since 2019, 2020, where we utilize them to help secure our workforce in COVID was a great investment that we made there to be able to secure a workforce in a very short time frame. I am responsible for incident response, threat intelligence and security engineering. So really the services that help keep Eaton safe.

Thank you. I'm StafaKeve. I'm the Chief Security Officer over at UKG. I've been there about 4 years now. Zscaler has been 1 of the strategic partners that have always leveraged and I've used it not only this company, but the prior company, where I spent about 8 years at brings. So when a customer is scale for over a decade now. So I think 2015, is when we started using Zscaler. So it's been in good time.

I'll start with Vopra. So you heard some of the innovations we have been talking about. And -- you guys have been a great sounding board for us. We build all the products in deep partnership with you guys. So going around again, what are some of the most interesting innovations that you see that announced today and which shift favor?

Absolutely. And I think it's been -- I'm always -- it's been always interesting to see the innovation that Zscaler is looking at either through acquisition or just organic growth building internally. So the SPLX, I think, is a very important piece because I was also a customer at SPLX pie to acquisition. So seeing that blend in and why we actually win that route and making sure that we're able to test our products because Zscaler been consumer -- a customer that provides AI software and software to customers being able to test those and validator is really key for us in terms of how we really look at it. I'm super interested about the AI piece because I think it changes the game, combining data, the AI graph and the access. And I think that visibility is a gap in the market that I think it's huge. Jason? .

I think the AI piece is the big area with the way companies are pushing AI to use it to make it work in your environment, the expansion of how they have visibility into what not only your employees are doing with AI, like your third parties are doing with AI and even what you're developing. I think that's a potential game changer there to get visibility across all those environments. .

Yes. All beating a dead horse here, but AI is everything for us for now, right? And I think the AI observability -- what we're talking about managing risk, right? We talk about shadow IT, right? It's really shadow AI. I can't secure -- I can't put a control I cannot govern. I cannot enforce policy if I don't know what's going on. So super important to me. The I will just say the maturity is so fast, right? Like we talked about 18 months ago and some of the things that we were doing compared to with the solutions that are being offered today by Zscaler and -- as I said earlier, we stepped into this and are running through the implementations as we speak on some of these prompt things and things that are going on. It's a game changer. We have to have visibility to manage risk.

Talking about specific scenarios. Man, I'll start with you. As you said, you've been a long-term customer at these killer. I remember working with you 14 years ago. talking about the benefits of local breakout. This is nothing to do with Zero Trust Ranch, but MPLS backhaul from your thousands of retail stores you adopted ZIA with the primary benefit of not doing local backhaul of traffic and doing local breakout. So that architecture has been evolving now to the point where there are appliances that give you 0 trust within the branch as well. So how have you seen that evolution in 15 years? And how has Zscaler technology evolved for your expectation in that time? .

Yes. It's amazing. 2010 already in 2026. I feel like we've kind of grown up with Zscaler. You think about when we start with Zscaler in 2010, what problem are we solving backhaul traffic, right? How did we do that? We didn't want to spend more money, do all that. We have 16,000-plus branch offices across North America. So -- it's a lot of traffic being passed around. So we really sat down with Zscaler. And it started off as really you think about it, it's a point solution where Zscaler is not anywhere near it is today, right, the size and scale.

And I feel like we've really taken that same journey in growth with Zscaler. And you talk about original Internet, the URL, the protection inspection to really what turned into cloud, right? And then when you jump into the cloud area and we talk about when we remember working early on with CASB, DLP, what's going on and how we've matured into that next generation of what I call technology into the cloud from a simple Internet world that we lived in.

And I've seen that growth as we've gone with Zscaler. And what I think it's always important is what was a single solution is what I would call a strategic partner today for us at Edward Jones with Zscaler. And I think fundamentally, what you see is they're either 1 step ahead of us or we're pushing them to develop the next technology. And I think what you can see is to be a strategic partner, we needed to create that ecosystem with them and moved along quickly with them.

I think the #1 thing I talk about all the time is, does our vendor understand what business I'm in. I'm on a financial service business. It's about availability. I got a trust, right? I need to understand I got to answer to regulators. All of those technologies and solutions as we partner with them and they deliver the solutions to us, Edward Jones.

It was -- it's really transformational as you look at it. And it talks about reducing complexity. And I think that's 1 of the biggest things we get from Zscalers, really getting in the middle, providing that 0 trust and being able to get in the middle and be able to do what we need to do, govern, pull policies in place, enable the business, securely. And I think when you look at where we've come and now we're into what I think is the next generation, not last generation, maybe for me, we'll see.

But what we really talk about is just last one is -- so you've seen how we've gone from Internet to SaaS world to now AI, and we're talking about what are the security controls are going in. We -- again, it was just natural for us. We could have looked at a third party and we did. We always do. Can you meet the requirements. Are we already implemented in that space? Can I reduce complexity? Do I really want to bring in another third party into the conversation with me? Do I really want to support another operational system? .

The answer is no, I don't. But I do need to make sure that they meet our requirements. And if they meet the requirements and exceed and help us create a better performance financially, economically, and to meet the regulators -- it's been a great solution and a great partnership with Zscaler. So for 15 years, I believe we've built that journey and what today is truly a strategic partnership. .

Fascinating. I love working with you over the years. Just moving to you, as you said, you started your journey with us on -- during COVID securing your users. But as we started building our Zero Trust branch solution, right, 1 question that everyone asks us when they start their journey, how is it different from my SD-WAN, right? And with you, we started working on this concept of Zero Trust factories. You have multiple factories deployed with Zero Trust appliances now or segmentation inside and with Zero Trust everywhere. Now -- we are also replicating the same framework in Zero Trust bank branches, Zero Trust hospitals now, how did you internally build the justification for our ZeroTrust branch or factories compared to SD-WAN, which is easier to deploy sometimes are things that networking people understand well. So how did you build that internal mind share? .

I will tell you this, it was really based on making sure that the sites were secure. I mean as a manufacturing company, we can't have downtime similar to what Wayne was talking about, but even more our production and our plants need to keep running on the SD-WAN and to provide us with the security that we needed all the way down to the device.

So we are really looking at as we deploy this and put it -- I think we're probably about 100 factories in to really make sure not only are we securing it at the network level, but at the device level and making sure that everybody has the right access in the environment. And I think we're on this 2-year journey now with Zscaler. They have been a really great strategic partner throughout this entire process.

We've been learning together, and we've been able to really make great headway when it comes to securing our plants to where we feel very, very much comfortable with. If something does happen, we're able to isolate it and secure it moving forward.

All right? I have discussed similar ideas with you as you said you had a returning customer. Your previous company had the same side of assets, which we discussed about securing with this Zero Trust branch architecture. But Shifting gears into your current company, you actually have been using a couple of our acquisitions, as you mentioned, both SPLX, you are a customer, symmetry systems, you were a customer with them as well.

And you have provided your input feedback as we were doing validation and diligence on these companies on why you like these companies. And one thing I remember you send an e-mail to Jay and I saying, you guys are on the right path with some of the acquisitions that you're making and connecting the dots. I would love for you to tell us how you articulated that story that you shared with us.

Absolutely. No, thank you. SP-9 And when I think -- when I see some of the things that were happening, and I think as a customer of 1 SPLX, and I'll give you an example of -- we're building software. We're testing the AI agents. We want to give it to 80,000 customers of UKG. And what we wanted to be able to do is have a fully automated testing capability that test some of our AI agents.

And this is not just and testing. This is we want to do different types of tests. So we want to do sentiment test. We want to do validation test. We want to do the security test. That's a key component, right, is in terms of how do you support for that. But the second thing is what are we actually protecting and we're protecting data, and we want to make sure that we understand identity -- and this is where Symmetry came into play in terms of as when we bought symmetry in Us, how do we make sure we connect this together, right, having access to the data, who has access to the data. what are they doing and what actions are they taking? So combining those things give me that visibility.

The third thing was if you tie that to what Zscaler does, in where it sits, the visibility and combining those 3 now you just have the full visibility of an end-to-end product stack. And that's why I said, you guys are on the right track by connecting these things together. Because not only you have an agent that sees traffic going from the endpoint to the Internet, you also have the visibility at the browser level.

You have the DLP that talks about policies that changes, "Hey, who's -- who can do this and who could not do this." But then what can -- what data can they touch? Can it just my payroll information? Do they have the right to touch that information. So if you combine all those things together, you just create a whole different game. Now I'm super excited about the developer side because I think that's a whole different game. If somebody is writing code in Claude code or you're using GPT or whatever Codex, that policy, the single policy agent is just powerful.

Now I don't have to go look for another product. And I think that's the key. I want simplicity but a platform, less platform, 1 or 2 that I can build my security around it.

Got it. No, this is very insightful. A couple of questions. I know we have 5 or 6 minutes left before we pass it on to the audience. in, you are in financial services. Frontier Labs came up with models that are finding vulnerabilities at piping speed. They are looking at how new attack chains are created, creating like finding a lot of things that were exposed out there, but not saying how badly they are exposed and how they can be exploited. We believe Zero Trust is the right way to start those exposures from being visible. So how is your security approach change in light of these front-end models doing what they're doing in recent time? .

Yes, it's crazy how fast things are moving out. I couldn't agree more. So it starts with trying to hide the attack surface, number one, right? So let's not look at that. But the reality is we're all trying to patch. We're all trying to do vulnerability management at crazy amounts of it. I think what you're finding out really quick from all of these new models that are coming out of exposing these vulnerabilities is 2 things where you used to be able to just focus on critical and highs.

It's not only creating new it's taking what we would call medium and lows, and it's starting to patch these things together and move very, very fast. So I think the reality for us is we -- everybody in the words of patching is we're going to have to automate more, right? It's just inevitable, right? You're going to have to have machine learning versus machine learning AI versus AI. It can't be human versus AI anymore.

We're not going to be able to keep up, right? It's just not possible. And then when we look at in the financial services industry, I can't express enough that with all the different regulators that are coming in, they're challenging us. They're already asking these questions. I look, any one of these Nissocks, take your pick from out there in the world, we're being challenged constantly is how are you handling this? And of course, they know about all this.

So they're asking the tough questions, right? They've increased the number of questions in the compliance space around AI. How are you controlling and how is it exposed? But more specifically, the focus is changing from a regulated perspective. And without these type of models and things that we're getting and really trying to say, block the attack surface, understand what the visibility is we need to prioritize, right? We need to find out what that is. We need to be able to fix it, not all can be fixed, so then there's the mitigating control. So it's fast paced. We just we're going to have to do what we can do to protect and prioritize.

Mustafa, stuff from your side? When you think about the scale of remediation that's going to happen and even today with the existing things that are happening, you need to buy time. So -- and I call it, being able to have segmentations in using the Zero Trust model to really isolate most critical and then focus on the most important thing in giving you the time. And I think that's super key in every organization. Anyone can't fix all of the problems you're going to have, so you need to be able to prioritize -- but I think with Zscaler that gives you that capability to actually isolate your network, segment, the critical areas and segment areas that I just use is out there so that you can maintain and have a better understanding.

We've been talking about this in UK like how do we think about the camera system? How do we think about the office, the conference room devices that are in our network, like this is huge and -- if you need to patch all of that in a day, what are you going to do? So these are the components of taking stock to really reduce the attack surface.

All right. My last question, I'll start with you, Jason. Is on how you are thinking about the AI security spending? Is this coming from your existing budget or you're reallocating budget for securing AI as new use cases emerge. How are you building this justification and the security budgeting insight? .

Yes. I think it's a combination of the growth of AI in a company like ours, we have to put thing forward, you have to put budget and to secure it as well, but we're also getting the incremental spend in cyber as well because the company overall understands the importance of this and be able to deliver on a secure AI environment. .

In? I would agree. Right now, it's not cutting anything. It's an addition to the budget at this point in time. We know cybersecurity, AI is something new and that we're adding to the budget at this point. .

I think it's a combination of both, it's reallocation and then finding the right investment for additional security.

Now these are great insights. And actually, on lease color side, what we are seeing is while we work with the cyber practitioners like yourself, we're also engaging more and more with chief technology officers, organizations who are building apps and they are saying, for example, I want to embed rec teaming more in -- on the shape left where developers are building apps. And -- we are working with this new emerging role of GFI officer, which sometimes is in data world, sometime in AI-specific role where they are looking at the whole lens of how they are enabling AI and new budgets are treated. -- because as you create more budget for AI, you need to secure that AI as well?

For example, SBIX is inside our product from the enterprise. So yes, it's a mix of both. And sometimes you have to do that.

Great insights. Thanks for sharing your journey and your insights with us. We have about 16 minutes left. So I'll open it for the audience here to take questions. We'll start with you. I'll randomly pick. I'll go across the room.

Thank you. This was all really good -- I actually have like 3 questions. I'm going to ask the 1 on the last thing you just said. So I think all of -- well, actually, Jason and Mustafa said that AI security budget was going to be at least partially come from reallocation from other areas. I'm just curious, what are those other areas of the parts of traditional security that you can take from -- like what's most kind of like that risk? .

It's actually not coming from security. It's coming from the business. because they understand the value that security is going to provide to them to keep the AI that they're creating safe. So it's not -- we're not taking anything away from IT or security itself. It's just additional funding that's coming in from the business. .

Mustafa?

So just curious Yes. For us, I think it's the some reinvestment in terms of areas when we think about all of our security stock, what security stack do we have that's below in lack of controls or it's not actually giving us the control we want. So this is something we evaluate annually and say, if we have a better if we can increase or security controls on AI stack because it's the most imminent, so we need to move some of those. It could be -- we are doing pen testing, for example, in this case, can that be allocated for prevention control instead of just testing. So some of those things are the things that we're evaluating.

We have the question in the front here. I'll move here. I think on the front here.

Can I ask you to please state your name and company name before you ask your question. .

Sure. No problem. Keith Bachman from Bank of Montreal. Very insightful. As we listen to customers and global SIs, a frequent conversation or identification of problem statement is understanding where the agents are, who owes what are the risk exposures. And a lot of companies come to talk about a value proposition associated with solving that problem statement, not just Zscaler but a number of companies. And so I'm interested from your perspective, when you think about that problem statement, which was identified here tonight -- or today, excuse me, -- is it one company you think you'll work with? Or is there more than 1 organization that will serve as that orchestration layer for lack of a better word?

And more broadly, are there other -- this is a Zscaler events, so unfair question, but are there other vendors if there -- that you think might be able to contribute to helping with this problem? .

Is that you want to take that one? I can start. I have a lot of agents. So when you think about the environmental ecosystem in terms of agents, there's a agents that you're building for internal use or agents that you may be building for your customers. So there's 2 components of that. Where Zscaler provides context is where it's sitting because you have -- it's sitting on the endpoint, it has the visibility as the network traffic.

So that visibility is going to be there that gives you that context. Even if you're using additional models, so you're using different or agents that are not specifically enterprise use, you tend to have that visibility. So the expansion they're doing allow us to see more and more. Now they are places where you may add additional context of additional products, and I think Zscaler is thinking about that as its scale.

But from my visibility today and what I see, and I think the more some of these companies become platforms and give you more visibility you would use them to actually give you information you need.

Another one. I think that's exactly right, right? I think the visibility in the single platform is really beneficial to us. I couldn't tell you that there isn't going to be a best of breed from another product or solution. The key there is when that happened is the integration level, right? It's really how a Zscaler or how is the product ex playing together and how do I integrate those together. It creates more opportunity.

But the real question is, if you're willing to go outside of the ecosystem, how much value does it provide to us, right? And if it really is that much value, then it's about the integration for us.

All right. We have a question somewhere in the middle. We'll probably go there. .

Eric Heath with KeyBanc. Thanks for all of you being here. I'm sure we haven't really touched on it, but more the SecOps side of things. I'm sure you all have other vendors that are using various SecOps organization, your SIM, our EDR, et cetera. How do you think about Zscaler as a partner in the SecOps Arena side of things with Red Canary and some of their ambitions there? .

I don't think I would be the one to take that someone else can take this one.

I'll take it. using that. Yes. It's very interesting. And I think the industry need a different view on operations and SIM. That's a very challenging market right now. Most organizations, I think, are struggling with the data coming in and then the remediation or time to remediation.

So I think from a product perspective, when I saw it, I think it's really interesting and he could solve a lot of problems that organizations have, especially the remediation piece. What was interesting to me is the data aggregation and how much they're bringing all of that data because that's -- today, it's very costly for most organization, ingesting that amount of data. Zscaler is able to do it effectively. And I think it could be a great product for many organizations in Houston.

Steve Koenig, Macquarie. Thanks for doing this. I appreciate it. So with you all cited AI as being the newest thing that's challenging year. With like the agents that Anthropic is offering, okay, being deployable on the desktop or on the endpoint or being deployable in the cloud. And potentially in the future, the big LLM providers like pinning certificates to that stuff and decrypting the traffic, so the Zscaler wouldn't be able to see it potentially.

And maybe I'm simplifying this too much, but with all this stuff changing so rapidly, like how mature are the solutions being offered today for you all in terms of being able to empower your employees but protect the use of this Agentic technology in cloud and the desktop clot in the cloud. How do you think about doing that? And is it slowing down your rollout of like these these agents, say, co-work or cloud, et cetera. Are you worried about that? And where do you start to protect.

So it's a really good question, and I think it will go back to partnering with a company like Zscaler to work through that. You're sometimes always playing catch up and you only could do what you can, but you have to really work with your security partners and your strategic partners in the space to be able to do that. And you talked a lot about -- you might not be able to see all of the traffic or everything that's going on, but it's really being able to see some of it, understanding your environment, getting that somewhat of a visibility to be able to take corrective actions in your environment.

So it's going to continually evolve making a strategic partner like Zscaler, providing them feedback, you working together It is really going to help the product grow to be able to get what you need to get done. .

I think it's easier when you have a scale Anthropic and Zscaler and an open AI having that integration because there is the power of -- for us to be able to to deploy, right? So today, I don't have the visibility that I want by giving my employees GPT and Claude code in my infrastructure because I need additional security controls just lacking from this morning, looks like that's coming. So that's amazing. So those are the kind of things that I think we need to be able to perform some of those things. And I think that partnership will allow you to say, even if there's a certificate change in the middle, later on, that partnership allow us to actually close that gap.

I think I just want to add 1 point to it. Yes, 100%. Look, as these model providers are becoming more enterprise deployed. -- they know that security is top of mind for enterprises. So we have partnerships. We have API integrations and expansion of our footprint on the endpoint with products that we have launched and what we are doing in public cloud. we are bringing that coverage to make sure there are no gaps left anywhere. But it is evolving landscape, and we are very focused on that.

Next question I think in the front here .

Brad Zelnick with Deutsche Bank. Really appreciate you all making time and sharing your insights with us. As pricing models across cyber and IT in general, evolve to more closely align and cover token costs and align the value what you're paying with the value that you're realizing, how do you manage and kind of mitigate and have visibility to where your CFOs aren't kind of choking you out and -- and where does Zscaler fit within all that? .

Yes. Great question because we just went into the AI solution. And I mean those are the control right, the capacity, that's the economy side of it, and it's probes and number of tokens that we're working with. And then I think it was interesting. I read an article not too long ago that was talking about people and companies are driving AI and AIU, they were measuring who was doing the most, but they were just using AI. They weren't being productive with AI.

So you have to change your measure as a business to what is the AI value providing, not just, hey, I used AI a whole bunch. But really, what was the productivity out of that AI and somebody was trying to win a contest to say, I hit so many things. So when you think about like us, it's -- right now, we do want to see people using it, right? You want that experimentation, you want that, and you're going to see some increased costs.

But the question ultimately is going to come down to, you're going to have priorities it's not this unlimited bucket of money that we all have, right? So when I look at probes and applications, I'm going to start to have to start to look at prioritizing my applications and understanding what I want to scan, what I don't want to as we go through that mechanism.

So I think from a business perspective, like anything else, the scale and the cost is going to go up, but how do you measure the value of the cost that's going up? And what are you actually running. You think about cloud cost when everybody just threw up things in the cloud and didn't manage any of it and somebody got a big bill at the end of the month, right? So you really have to sit down and start looking at from a business perspective and say, what are you allocating -- what are you permitting and put some controls and access around this, so that you can see the difference in what the money is being spent on?

We have 3, 1 in the middle and now. Let's start there.

Okay, Jason. This is for Jason, Mustafa. Ashish Bhandari from Through line Capital. Both of you spoke about using your Zscaler deployments to have better visibility into developer usage. That's an interesting use case and not something I explicitly thought about before.

Maybe can you double-click into that? We've seen all the cogen tools go pretty nuts over the last 12 months. So I'd be curious how you're using Zscaler and other vendors to address that.

So there's the capability -- Zscaler because it sits on the endpoint, you start to look at the capabilities to talk about, which is IDs -- so the engines that the developers are using to write code. We have seen a lot of supply chain security happening lately.

And this is the visibility. How do you use that to actually get better traction and make sure you know what the developers have. The second important piece is the API integration they have into the AI agent that allows you to see what's happened and what the developers are doing from a policy standpoint. And I think with some of the enhancement that it's coming, you can even have predefined policies and say, someone can do this all clear the parameters that you can actually do.

So I think that's powerful, that's missing today. And if most organizations, you either have to build something or go find new provider that's doing that. So that's, I think, in my view, it's --

And just to add to that, like look, from our side, we have some very large customers who are big technology shops. We are big developer populations, right? So even with core products like ZIA, even if you don't think agents in AI security, we have been covering them for many years. So 100% a big focus area. I think there was a question you had in the middle.

Meta Marshall, Morgan Stanley. Maybe a couple of follow-up questions. Jason and Wayne, you guys kind of didn't talk as much about the Agentic SoC. And so just kind of wondering what solutions you are using to kind of manage a lot more data kind of coming in and a lot more signals that you guys are getting? And then on the second question, kind of on the final point around pricing, it sounded like kind of the response to Brad's question was, we'll get the AI costs under control, and that will kind of level set the rest of the cost. I guess, just is there a comfort right now with kind of token-based pricing within security that kind of mirrors that AI pricing.

I'll take the second.

I'll take the first one.

There I'll sit on the backside on the pricing. Are we comfortable? I think we're learning what Comfort looks like. I think you're right, it's pretty new to us. So where is that? I know there's going to be a ramp-up, Absolutely, right? We we're going to purchase so much, especially on the probe side and we're going to go out there and we're going to start to see the value of it.

And naturally, the organizations are going to continue to grow. So I think there's always going to be a capacity increase with that budgeting. I think he'd be hard-pressed today. It's easy for us to go out there and say, evaluate what we have today and say, this is how much we need. I think what will be interesting is when we get 6 months or a year down the road and we start to see where this tiering level goes, does it ever level off? Or does it grow at a certain pace.

Comfort, I don't know. I think there's a strong word. I think there's an expected cost -- but comfort is something I think we just have to get better at. I think like I always use the cloud, like we just like to throw everything out there, and we'll see what happens. And then everybody started to figure out how to manage it. I think that's what we're going to learn very quickly is how do you manage AI expenses and growth? I think comfort is probably a strong word.

Yes. I'm not going to talk to vendors and I can talk to strategy. Yes, as an organization, we are definitely looking at how we can use a Agentic SOC. So because as you've heard, you're not going to be able to keep up with human speed anymore. You need to use genetic AI to be able to help you. .

And that's a big focus area for us. You will hear that tomorrow's keynote, which is going into what we are doing in the Agentic SOC. We have always integrated SOC source and are 1 of the highest fidelity security data provider. But our data is unique, and we can build a lot of findings on top of it and by integrating third parties and some of our investments and have a lot in, you will hear what we are doing in that space.

We have 1 last question that we can take. We'll go in the back there, and then we'll wrap up.

Yes. Peter Levine, Evercore. We're at the Zscaler conference, but maybe if you take a step back, if you think about identity, network, endpoint, cloud security -- what's the first layer of defense that you're defending now against some of these AI attacks? I know you're investing a lot more in Zscaler and their AI products.

But if you think about identity, your endpoint, like where are you spending most of your capital today to the founding business.

Identity still to me is extremely high. I mean you start with identity in the roles and segmentation as you talk about that. And -- but it all bleed right back into Zero Trust where it happens, right? So whether I have identity, and I have ZPA, right, and I can apply an individual to an application on our system and really put that enforcement policy. But I feel like it always starts with me in the identity in the credential space because once I know that, then I can control and enforce policy through whatever mechanism I feel that is, whether they're coming through an Agentic either not a real person or if they're a real person or not, then I apply that. So I think identity has to be a strong focus and then everything else builds from there.

I would say identity first, data second, identity, you use it to make sure you know who's coming in and what they need access to. In the event of zero-day or they just work in and get access to the data. It's copying that data and validating that. So if you connect those together, and I think you have a good chain of security, which I think Zscalar is trying to get up.

All right. I think with this, we'll wrap up the panel. Thanks for your questions, and thanks for sharing your insight gentlemen. And next session, we'll need a few minutes to set up the stage. So give us a couple of minutes and we'll get back.

Yes, please. Okay. We're going to get ready with -- we have plenty of time for Q&A with all these folks. So please raise your hand. We have some mic runners -- maybe start up here with Brad. .

Thank you again. Can you guys hear me? .

Okay, Brad Zelnick, Deutsche Bank. Great to see you all. Another Zenith Live in the books. Great stuff. Mike, I wanted to direct my question to you. One of the surprises coming away from Q3 results was guidance that we heard, which was incrementally conservative -- and the context around a few key sales departures as a reason for that, which in the context of a company with 8,000 some-odd employees was just a little bit surprising. And not to dwell too much on that.

But looking forward, can you just talk about the resilience of the go-to-market organization, why the pipelines and the relationships or institutional relationships and the risk that we bear going forward? Or is there going -- because we've all as investors seeing these movies is there risk of fallout that you've got dozens and dozens of others that are on their way out the door? Or just any help you can share with that would be great.

Yes. Good question. So I think the unique thing about that was just -- it was just 2 at the same time. That was it. right? Normally, I mean there's always going to be turnover, especially in the world that we live in today with AI. There's always this new hot company that people want to go to, and they've got formal. A lot of the people here, and I always get so energized, I come to this event, -- and I hear the executives talk about how much they love their account teams.

That relationship is so important. We have so many talented people here that love it. They love what they're selling. They love our solution. They love the future at Zscaler. So I feel very confident that our strongest people are going to be successful and continue to thrive here.

They also want do they have a career path. And so as we grow we put a lot of time and effort into career passing these folks to make sure we keep the right people on board. But it's healthy to have some level of turnover. And in both instances, the people that left. One was a mutual thing. The other one was a little bit more -- maybe a little bit more of a surprise. But we've got great people on the bench to backfill. So it also raises the game of other people. and helps on that career pathing side. Do you know what I mean? So they see a future other people when 1 person goes, somebody else gets promoted, maybe you bring in some new folks -- but a lot of times, you have a strong bench, and you can promote people. And that means there's another set of promotions. I got under that for worthy people. So it's an opportunity as well. .

Yes. Okay. How about John, over here. sit outside the side? .

It's John Difucci from Guggenheim. I have a couple of questions. I'm going to come to Mike, too, because I actually never met you and I have always wanted to. So we hear a lot about a changing go-to-market strategy when you came on board. -- become a much more strategic partner with your customers, and it all kind of makes sense.

And I think your product people have really built up the platform so that it's gotten to be more of a platform rather than being used for a couple of products. And we do hear about a ton of large deals in the pipeline that continue to get pushed out. And I'm just curious like I'm probably not the only one, everybody checks here into the field talks to partners. I just wondered like what's going on with that? Is it -- are people sort of waiting to sign large deals, they are they fully committed to Zscaler as a platform? Like I heard your customers up here talking about Zscaler and solving the problem of AI, which I don't think any 1 vendor does that.

But I don't know, maybe you guys think you do by yourself. But how would you kind of talk to that topic. And I know there's a question in there somewhere, so I apologize. But there's these big deals. You're a more strategic partner to your customers, but they don't seem to be closing as much, I guess. .

Yes. Maybe do some deals happen faster than expected, too. So it's kind of a balance. But the deals that take longer to get done, it's usually because there's a lot of testing that they have to do. And there's a lot of people -- it's kind of a political landscape, a lot of people you have to get on board. -- from different groups, right? So that takes time. They want to go through the testing. They want to see what we can deliver.

Sometimes there's new requirements they want to see us deliver before they're ready to take that deal to the next level. And I think that's 1 of the strong suits of Zscalers how closely tied we are to customers, how we listen and actually, we deliver on those requirements faster than the competition. That's what I hear at least.

I'm probably biased, but that's what they tell me. But sometimes it takes longer, right, to build in all those requirements. And then they're going to say, okay, now I'm ready to go because I've done the testing. Large companies have -- they spend a lot of time doing testing. -- a lot of time doing testing. Hopefully, AI can help solve some of that problem, right? And they can speed that up. But we can plenty of deals that happen ahead of schedule as well. So I'm not super concerned on the time it's taking to get these deals done. That doesn't keep me up at night. .

Maybe I can give a perspective of a CIO because I was a professional CIO before getting on to Zscaler payroll. What I think Dhawal mentioned this earlier as well, See, the large enterprises are also struggling with these AI Sunami coming in. The -- who is the owner inside large enterprise to drive -- some people are pointing chief officers and some people make the data person, be there, some people keep the infrastructure.

So -- but what's happening is they're all realizing that the right to play and right to win to protecting AI is the network providers, right? And then among the network providers, so we are getting a lot of traction around how our network teams and the are bringing the application teams and the officers, and that's where some of these testing cycles will take longer. But we see the momentum. I see the recognition by some of the CIOs that Zero Trust providers have an advantage. And that's what Jay highlighted in the town hall today as well or in the keynote.

Keith up here in the front .

Keith Bachman, Bank of Montreal. Thank you very much. Jay, I wanted to direct this to you as you -- I don't know if you were listening to the last panel, but identity was no pun intended to identified as 1 of the key areas of spend as we look at the next period of months and probably years.

And if I think about some of your offerings -- it seems like you're encroaching on identity, right, in terms of your value proposition and Swami we were talking about this last night. But I'm trying to understand where does Zscaler's value proposition start and stop relative to the identity partners?

And are you frenemies are directly competing, and this is particularly related to areas such as governance of agents is what I'm referring to. Thank you.

This kind of builds upon the question that got asked and ask listen to the answer as well. To me, the question is not that is identity more important or network more more important. EDR does what it's supposed to do on the endpoint. It's like watching what's inside your house. It's useful, but it's contained to that. Identity is the starting point of access to something. Identity can be used to do old school access to put you on the network. Then identity is useless.

Because identity puts you own the network, you own the network, you're going, you can go all on them. Identity combined with Zero Trust together is the real solution that says only this entity can talk to that entry. So our view has always been. I don't see tight integration. And then we on the switchboard that makes the right connection on one-to-one. So encroaching on identity.

Let's talk what that mean. The basic identity starts to bid John identity is this. We get that from Okta today or Microsoft. But then there are a number of things on top of that, which device is John coming from? We know that because traffic comes from, which location is it coming from? What's the behavior is the traffic growing less than us. So we're adding value on top of identity by looking at a bunch of attributes.

We call it additional authentication services we build on top of what we get the basic identity that we do that today for users. Now this scope will become more important for agents because agents need to know a lot more than basic identity. The question is, should Zscaler provide the basic identity or what should it provide?

Our view is that every provider that's going to allow you to create agents, Microsoft, AWS, Google of the world. Identity just gets created. The basic identity who is -- what this agent is comes from that agent. I don't need to compete to get that identity. I take that, I become the switchman, but now I can add a number of authorization services on top of that, skills, tools, access, all the other stuff.

We do all of that. That's becoming extremely important along with that. So yes, we are not directly competing in the basics of identity, but we are competing to deliver solutions. Customers don't buy identity for the sake for identity, customers why identity or ask to access on application services. And now the question you're going to ask is Symmetry identity. Yes, no. It's the graph kind of say, who is tango. It gives us meaningful information -- so we may not do the basic identity to compete, but all the other value to make decisions about what to connect is very important for us, and that's really what our focus is.

Maybe I'll just add quickly. It's we will and will continue to use that as context. It's very important context as well as the authorization of what that identity is allowed to do. So that's part of the information we use in making our policy decisions being the 1 that grants that initially, that's a whole structure around who the business owner is of that.

It gets very distributed in organizations. It has to cross lots of different parts. I don't know that there's a spot where we have -- like where it's critical that we own that piece. We need to know it, and we need to keep up to date because the other part 2, when you look at this intersection is being an in-line solution or being on the endpoint where the action is happening, which were in both of those. There's also a whole question of that initial authentication or authorization that's granted to the application what happens if the behavior or pattern changes while that session is open. We're in the best spot to actually take a real-time dynamic action, say, based on new information, I know, I'm actually going to end that session or that conversation.

And that's a great spot to be able to enforce. And we do that in a couple of different areas and that whole discussion is coming up on agents as well, too, right, in the middle of sessions. Like what is it the risk changes or the posture changes already authorized the sessions there. Who knows what to do -- who's in the spot to take that action, we're actually in a very good spot to do that. .

So I may add 1 more comment to clarify I get asked by many CIOs. I said I thought identity provides policy of who access is what? How come Zscaler providing police okay? -- right? They get confused. My simple answer is, when I go to an international airport, they scan my driver's license or my passport. And Passport that computer makes a call to a database of passports. Is Jay's passport valid or not. That's identity. But I'm somebody to sit in line to allow me to go or not go. .

We are in line in space have thing. Identity once you get checked out identity is out of the way. Identity may have groups to say who can do what, but then it's out of the way. Being in line, we're able to add additional value, authorization, behavior and all is what we can do. That's why we play a very important role. That's why having the best identity for us is not that critical.

Let's go ahead, Catharine.

Catharine Trebnick, Rosenblatt. Can you unpack why you said ZIA and ZPA are growing? Is that due to the acceleration of ethos in the landscape. But just you picked that up in your opening remarks. .

So when you talk about ZPA. So there's Zero trust part for CIDP users, then this for workloads as well, and we'll take the same ZPA for agents as well because at the end of the day, -- the goal is who can access what application we are having. I think. So on the ZIA side, largely, that stuff comes from new logo acquisition because most of the time, when they buy a DI, they do it for all users because they must protect all users. ZPA, many times, they only bought partial users because it started out by replacing VPN. Now under metals.

They're basically saying every user must be untrusted. So we're seeing a lot of interest for people who have bought ZIA, but haven't bought ZPA so far or who have bought partial EPA want to go to full ZPA. Those areas are directly beneficial because that does 2 things with ZPA, you're hiding your private applications behind us. Also with ZPA, the lateral movement goes away.

So we see Methos as tailwinds for it.

Second row. A couple of questions right there.

Roger Boyd with UBS. Jay, can you compare the level of urgency you're hearing from CISOs today to what you saw during COVID? And I think you laid out very clearly why Zero Trust architecture makes sense for this environment. But I think the other question is how quickly can customers get there. And in COVID, we saw sales cycles meaningfully compressed. I'd just love to get your perspective on what you're hearing today.

It's a very good question. The urgency for Methos is actually higher in many ways. I didn't see the board level discussion happening. -- as much with COVID. -- they wanted people to come back to work. But almost every CIO I have talked to us so they said, we've got a task force, we are reporting the Board every week or every 2 weeks on the progress we make. It's that level of stuff happens. The difference is the following. With COVID, you went home on Friday, you needed to access your work on Monday morning from a home.

So the urgency was give me something to get started. With Methos tell us, they're struggling there figuring what do I need to do? The first thing they all wonder is, what is this Methos, what does it mean to me? And the way Anthropic has done it, it's kind of a mystery thing out there. You don't know, you wonder about it.

But as we had talked to the customers and explain to them, they're looking for practical steps and saying, what can I do and report to the Board that I had done AB and C and I'm making progress. So all of our discussions have essentially led to essentially deliverables where you ask that they want to work on fixing volabilities. But -- our current customers are actually working on hiring their applications behind us.

They're working on moving to Zero Trust everywhere. They are looking at the users not being on the network and branch projects are actually gaining more interest. But I think it will take some time. I expect the momentum for ZPA kind of stuff will happen faster in the customer base.

The new logo takes a little bit more time and testing. But I clearly see the interest in moving more towards Zero Trust with Methos than it was before Methos.

This is Shrenik Kotari from Baird. So Jay, you have talked about how agenting exchange could be the largest transaction-based profit-based monetization opportunity. And today, sort of double down, talk about the Agentic transactions or order of magnitude potential Zero. We heard from the customer, it seems like from an AI security and a Agentic prioritization, they are adding to the budgets, there's appetite.

In terms of just the core monetization meters, right, you announced AI broker. So I'd love to hear more about how that becomes a commercial manifestation. It seems Customers are anchoring towards agent identities orient first. You talked a lot about identities being sort of the centerpiece. So just curious how are you thinking about this monetization strategy.

Right? So you saw a genetic exchange and the traffic that's coming through exchange for agents, essentially based on traffic or call it, number of requests, which translates to tokens becomes the commercial mechanism for us to monetize for it.

And -- we just launched it. I think -- so we are seeing a lot of interest building our customers who work with us early stage, POCs and all that kind of stuff. I can tell you, I'm not seeing so much interest in any product than exchange for agents and being a critical product like this.

For example, on contrast to the area. Asset Management, do they care about it they do. They don't like it. The right time in the do. But they know that Agentic is a hard and important problem to solve. They're working with us very closely. The exact pricing and all, we are still figuring out the way we do pricing I work with a first dozen, 2 dozen customers, figure out the traffic flow pricing that needs to be done. So pricing will probably get firmed up in the next couple of months as we see traffic, how much needs to be done, but I see lots of interest. And I'm looking forward to see the growth, and we'll share with you as we make progress in this area. But it's an exciting, challenging problem -- the #1 reason I hear from CIOs is why we're not able to roll out these Agentic projects in production at mass scale is lack of governance and data security issues.

Let's go over here to the left side my left. .

Thank you. Gray Powell with BTIG. So thanks for hosting the event today and a good presentation. So I understand that there's a lot of urgency and discussions created by Mythos. And I'm just trying to figure out how that materializes into demand. And specifically, do you see it driving more interest in the existing, like the proven products such as ZPA? And I'm asking because just the marketing on a lot of the AI security products, I just have to -- it sounds the same. It's confusing to me. I think it's confusing to buyers. So I'm really interested if you could just talk about what you're seeing from the perspective of core product demand versus new AI security products and just how you see that playing out in discussions? .

I can start and Swamy, you can add to it since you're very heavily involved. .

When we -- I'll give you a detailed answer. When we do our meetings with customers about what can I do to protect against metals. We have very specific 6 recommendations, and they have become as a result of lots of discussions. -- one, higher attack surface. That's where ZPA plays a very important role because you're hiding attack surface of your private applications, okay?

Number two, if you get breached, how do you make sure the breach doesn't spread around? It needs Zero Trust at the user level, fast because use the weakest link. That drives demand for ZIA, ZPA bolt. And then we say, make each branch like an island that make sure that the infection doesn't spectrum branches. So it really builds demands are bold. -- users, branches and even workloads.

Number three, if you already got ZIA/ZPA deployed, which are foundation for your users, you need to make sure they're properly configured. We are doing validation of the configurations to make sure it's the best practice of forward configuration.

Number four, while I assets are now directly linked to metros. They're just showing up in every enterprise, and those AI assets are creating risk. So understanding what you have with the risk is #4. Number five, you should discover and fix will it prioritizes and fix vulnerabilities.

That's everyone is expected to do that. Six, you change from doing right teaming a couple of times a year to continuous automated that teaming that's driving demand for our red-teaming products. So this is the list of recommendations that go through and customers prioritize it, but they realize that Zero Trust becomes a foundation to do more and more of Agentic stuff. And that's obviously the demand being driven for AI as well as Zero Trust users and other things as well.

Yes. I would say that if you look at the discovery that we announced, right, that is something that CIS want right now. I always tell my team that you get to build the products that customers want yesterday. So that is the demand on that and the POCs are going very well.

And those conversations quickly switch into how can I now manage it. And many of you have seen this NVIDIA 5 layer AI stack. So if you look at the top layer, the application, so we can protect it with secure and then the next 2 layers models and LM they would want to make sure that you govern that as well. So what's happening is when users were using Internet, you would type www. something.

In the era of agents, the equivalent of WWW is MCP. So the AI broker that we launched was effectively how to really govern the activity that's happening. And within MCP, you could renew skills, tools and other problems that we can put in, all of them have to be governed. And that's why people are excited beginning with the discovery.

Maybe just 1 other comment to add on that. I think the -- when you think about protecting your organization, the core capabilities that Zscaler has and as Jay mentioned, and specifically the ZPA having that application, that attack surface of your applications, that's that's the most obvious thing that everyone should be doing and pets driving a tremendous amount of conversations.

But it's also when you have that discussion, it's clear and it's understood. Now people need to do it, and there's the -- getting the mindset to make changes, but it's not a hard discussion with people on that. And I think what I've seen with Methos is that recognition of I know I need to do more. Now if I haven't done this already now, it's time to do that. That second part about MCP servers and all the assets that are part of using AI for productivity it's not a chicken and egg, but you're not going to spend all your money on securing something that you haven't even figured out how to use yet, right?

So all of those pieces are popping up in organizations. And so I get why you say it sounds confusing because each of those terminologies, those pieces of things that are part of using AI for productivity. It's it's cloud co-work, it's MCP servers, it's A2A, and it's all these different pieces -- it's at the endpoint, it's what am I using a foundational model in the cloud.

Everyone is trying to figure out -- how do I use it? How do I get productivity. And I think in the earlier discussion, people talked about, am I seeing the actual outcomes versus just usage. But the fast follow on that is, so if I'm really going to use this at scale in production, how to do I heck I secure it? And so everyone is looking for those attach points. And I think what you saw from double this morning and in the discussions today, we do think we have these right 3 pillars for how to look at that.

But that is all new for folks, right, in terms of where they go in to spend. So it does sound like a lot of people, like it's that gold rush type mentality, where every big company on start-up is going to say, "Well, I'm going to help you with that specific piece of it, right? And that's why you see a lot of common storylines in there and everyone saying, I'm going to be the one who's going to do that and we believe we can, too.

Dave I'll grab him over here on my right, your left.

Steve Koenig, Macquarie. This 1 is for Mike and if Jay wants to follow up, maybe he'll want to as well. So just maybe extending the last question when it comes to these conversations that become about securing genetic AI in the enterprise, which is complicated and difficult for for the enterprises to, number one, figure out what they want to do with the Agentic AI and then they got to secure it. So Zscaler has has an ability to enter into those conversations and you have solutions that help with that.

But I'm wondering, how does that affect your sales motions in the sense that what used to maybe be an easy conversation about ZIA or ZPA or even Zero Trust everywhere, now becomes a conversation that is potentially much more complex because the whole is year protecting AI enters into that. And then maybe does that change the nature of your sales cycles that would have been much easier. So how do we deal with that? I guess that's the question.

Well, it's still early, right? And we're learning. There's new personas that we're going to have to build relationships with. So kind of the way we've got the sales org set up -- we have Doble's team that has some core folks that are very, very knowledgeable, go very deep with those personas today. And then we're training people by region, by area where kind of the major buying centers are to make sure we have enough people that are able to have those Level 3, whatever, Level 2, Level 3 conversations.

But yes, it's going -- we're learning as we go. But there's 1 thing that's very clear. Everybody wants to have that conversation with us and they all do feel like, especially if they are a customer today, they feel like they would prefer if we had the right solution for them because we're already there in line, and we see all their traffic. So we have this advantage, and we feel like we have the right to go win they feel the same thing.

So if I may add, many times, the notion of who is the buyer matters Zscaler has traditionally sold to 2 most important buyers. CIO is #1. CISOs #2. And why it's a CIO #1. 6 years ago, I used to think that CISO was number one. Because the transformation is driven by the CIO, CIO, CISO and Head of Infrastructure Network and these folks play a role. A lot of the discussion gotten stuff, go to CIO. Some companies do have Chief AI Officer or Chief Data Officer, then actually into gets made by the CIO. If we didn't have access to the CIO or CISO. We would be wondering about which solutions take all the security solution we have. They all lead up to the CISO. Now AI solution lead up to CISO and CIO both.

I think from access point of view, fairly well covered. The key for us is how do we streamline our teams to be more effective. There are some similar things to what we have been selling there some different things. Take what's similar? The notion of exchange for agents user branches is pretty similar. Yes, there's some nuance things underneath. But explaining that philosophy of we did it for users. Now we do for agents. Is fairly easy to explain and get the stage going via it count exact, then we can pull in SCs and subject matter experts as they are at.

But also, we have evolved some of these specialty teams, too. I mean Mike did specialty teams in his previous company, where they actually have different buying centers, at HR in 1 case, IT in second case, some in other case. So we have a specialty sales team, for example, for data security, which can get pretty complicated. Now data security is still sold to the CISO someone under the CISO.

But the technology functionality can be complicated. So we had experts who actually talk about that area. Similarly, we had a few others. We also evolve when a new product comes in under product management, we learn, we understand and we figure out how to go forward with it. But for AI overall, everyone wants AI. So rather than having a small specialty team of AI, we actually want everyone to sell AI, that they'll be backed up by some experts, domain experts, nonspecialty salespeople, but domain experts who could be pulled in for deeper discussions.

We are learning the process that the part of selling AI Protect, which is asset management, secure access and the guardrail and red teaming is fairly straightforward. We learned quite a bit in the past few months. The learning will probably happen about our exchange for agents in the next couple of months, but there's a high degree of interest.

Todd, right in the middle?

Todd Weller with Stephens. A question for Adam. Adam, you come from the SecOps world. It's a new space for Zscaler not as known crowded space. What's the strategy for breaking into that market? And what is it about the solution that you think is differentiated and will resonate with the customers?

Sure. I think you heard a touch of it from the folks on the panel before. But yes, it is new. So they're also learning how we're approaching this. I think the biggest piece from a SecOps standpoint, right, is that organizations are and do have centers of gravity of data.

And so if you're a Zscaler customer, we are a center of gravity of data with what we do with ZIA, ZPA, DLP, what we have from our client perspective. And -- our approach to this is how do we bring that together in a meaningful way for the purpose of security detection, investigation and response, right? Because that -- we have detections, we have signals. We have contacts -- and historically, we haven't brought that together in any fashion to help a customer through that investigation process.

We said you can stream it somewhere else to do it, but we haven't offered that place. Yet we also had capabilities like our threat hunting services, where we were uniquely positioned to use our data to find incidents that otherwise would not be found by streaming it off to some other SAM or SecOps product. You wouldn't see it on the end point. And we had people who like that service from us, but it was a small, I'll call it, like pilot service.

But at the core of that is they were writing detections that could be found across that Zscaler ecosystem of data. That's part of what we've been bringing together foundationally, the data fabric that we acquired several years ago is what creates those entity relationships between all the information we have third-party data like identity, contention like vulnerability and exposures and asset information -- and so we wound up having this foundation, we did the Red Canary acquisition, which brought in meaningful detection libraries and skills around how do I run detections against not specifically Zscaler data, but any third party that you want to bring in.

And this past year, we've been bringing all of that together. What we're bringing to market though is the software platform. Because Red Canary was an MDR. So if you wanted that service you got a service. If you wanted a software platform, you got to service because that was the only thing that they had to sell.

And had we not acquired Red Canary part of their road map path was how do I deliver this in a more cost-effective software platform way, but that wasn't what they had built, and that's not where their 10 years of experience came from.

Zscaler, you know that's what we build, right? And so we had this path that we were going down -- we had the Red canary acquisition in this year has been about bringing that together where that Agentic SOC core platform becomes that foundation where I should be able to go at a minimum to any Zscaler customer and say, -- we already have this data. I can make better use of this for you in detection, investigation and response whether you send me anything else or not, happy to take more, and I can take more.

But even if you don't, I will show you how I'm using Agentic framework to go through each of those steps and help you investigate incidents in a way you could not do before -- and then if you want to send that off to your SAM or whatever other product, you can do that, too. And know, by the way, if you want a service on top of that, we have expert skills for both threat hunting and full MDR that we can now offer on top of that.

So that's what you'll hear tomorrow in some of the keynote and then in a more formal launch. And I think that's a valid entry into this space. it also balances long answer, but I'll just wrap this up. It also balances, I think, the long-term question about what will happen in the Sakani world where there's this continuous, I can do the job that could give me all the data. I don't need any of the data. I'll just send agents to go get all the data when I need it. right?

And everyone I speak to in the SecOps world, you need a foundation of that data. Like we're not at the spot where this is just real time. Agents are just going to go grab data from their source at the moment they need it, you need that initial core foundation. And I think organizations will have several, right? We won't be the only player that's there, but I think we can also live and it's important, we can live with those other players that are in place, which I think is a requirement for entering into this market when there are already other big players and crowded in there. So I have to be able to show a customer that it's okay that you're working with CrowdStrike who's a partner of ours, we can be in here not just, oh, the only way this works is if you only use us. .

Okay. The right side here in the middle on Tazz and then we'll go to Eric.

This is Tazz from ROTH Capital. I had a question. I had a clarification on the Zero Trust for agents. Is there dependency on customers having ZIA and ZPA users before they can use Zero Trust for agents or can customers who are completely need Zscaler also use the transfer agents.

They can go on, you don't have to buy the user before you do agents. If you have users, it becomes easier because you understand this. But this is not a dependency. .

Got it. And the second part of -- second part of the question is, you gave us the bookings number for AI protect $100 million over the last, I guess, 1 year. And my question was, again, how much of that is coming from net new customers to Zscaler versus upselling. And when customers buy AI protect existing customers, what is the typical uplift that you see in the deal value?

It's a mix is the answer. The uplift is a harder 1 we have. I haven't looked at that personally. I don't know if anyone here knows the answer. .

Yes, we'll have to stay tuned for that one. We'll come back to you on that. .

Can you pass it to Eric, please? So you're right. Awesome .

Eric Heath from KeyBanc. Jay, I guess this 1 is for you. I mean, tracking all the breaches we see, I think a very, very common shortcoming point of exposure is the hardware that sits on the perimeter, right, the firewalls, the SD-WAN to VPN, et cetera. And you always talk about how this is a weak point. And maybe I'm over extrapolating and maybe I'm reaching, but it seems like Mythos can only accelerate that short coming from the hardware vendors. .

And we know there's several hardware vendors that are constantly patching and being exploited and et cetera, and Methos those only accelerates that. And then the ability for these hardware vendors to support these large fleets of hardware devices that need to be patched and fixed and whatever just we're always talking -- now the conversation is the compressing time line of between vulnerability and breaching and exploitation. And I know your answer is going to be yes to this, but do you think that this accelerates the transformation from hardware to the Zero Trust Exchange. And like I said, I think your answer going to be yes. But like in practicality, do you think this is going to be a real accelerant for customers to think .

So having had probably well over 100 conversations CI and CSOs in the past couple of months. light bull for better understanding Zero Trust is going up. For example, we talked about users being untrusted should never be on the corporate network. Some of our very progressive customers have already done it. many have known they're saying, "Oh, I should be doing that. So the understanding and learning for Zero Trust is going up, understanding and learning that 300 branch offices, the firewall face in the Internet is exposed to the entrant. And that is not a good thing, is going up.

So I do believe that Methos is becoming an excellent or adoption of Zero Trust. And it's going to start differentiating the firewall guys who always talk about we go Zero Trust say because they see the difference. Now they ask a question. No, -- do I have a lateral moment or not, what's going on. So this is happening. And also the point you said, patching. The bigger boxes you got everywhere, the more software functionality, you sit in a box that's scattered around.

The bigger that is because bigger the issues out here. The less you got sitting on the device is less likely you have issues out there and more likely take the branch office. We do have a branch, -- we're trying not to get there, but we are there because the customer needs.

We run them, we manage them, we operate and we upgrade, okay? -- essentially. So we are taking that risk away. And also in the traditional world, there are firewalls sitting inside the campus deep inside that only under customers' control. Those are the kind of things sitting out there. And I do believe that it is accelerating this stuff. And hopefully, we'll show you results in coming quarters.

Okay. We're going to go over to George. .

George wane with Oppenheimer. Kevin, bringing you into this since you haven't had a chance to talk yet maybe giving us some perspective on -- there's a lot of opportunities here, how you're prioritizing your investment with respect to your product and sales and marketing. .

Yes. No, I appreciate the opportunity to answer the question. Some of it has been answered as we've kind of gone through this conversation, right? So as we think about when we bring new innovations to market, the way we're thinking about AI is different than we've thought about some of the other products. So we have a dedicated team with Dave and Swamy who are acting in effect like a startup within the organization to really move with speed. But then they're working directly with Mike's organization to enable as many of the sellers and the people within his word to be able to sell it broadly.

Obviously, AI is just a very important element, both in terms of what we're providing our customers, but also internally. So that is a priority internally if we think about in that regard. And then the other products, we have specialty teams where we specifically identify resources that can help Mike's team go and sell. And actually, those live within Mike's team. So they're part of them. But that's how we think about the trade-off in different investments. And we, like every other business, go through an annual process of setting our operating plan, and we identify priorities. And AI will continue to be very top on those priority lists. .

If I may add, while our portfolio has grown, has become pretty large. We actually say no too many projects and many initiatives. EDR has been asked for many times. We said no. Identity has been asked many times. We have said no. So we're fairly disciplined. When we do projects that are very synergistic to Zscaler. They don't require as big of an investment as it would be if we were to do the old way. 5 years ago, I told you when we did sandboxing. Literally the amount of effort that was needed to do sandbox on Zscaler ZIA platform. 20, 25% of the total effort if as compared to if we're done or an independent company because rest stuff is already in place. They're taking traffic, they're opening files, all the stuff has been done. I talked this morning, Agentic Xchange, building on top of zero-tu exchange. -- only 70% of the pieces that are there. 30% is what we're adding. Take, for example, you heard about Zscaler cellular.

It's a very cool and exciting area of opportunity. What are we doing in amount of after needed -- it's a fair, very small team that's leveraging all the back end, but a new use case to take traffic, take telemetry from these IoT devices. So being doing the smart thing being around our core competency is what makes us more productive in delivering more products with great returns. .

I think we have time for one more. Let's make that Rich in the back. .

Rich Poland, Wells Fargo. I think we talked a lot about just the investments in a lot of things, AI on the product side that can really I think, drive a lot of expansion with the existing base. But when we think about 1 of the things that was said last quarter, the new logo side, it seems like that's more of an emphasis now targeting that 200 to 100 employee range. I guess both Jay and Mike, can you talk a little bit about just what's needed there? What's needed to enable that? What did you see that prompted you to want to focus there more? And just kind of any color around what you're doing there. .

Mike, and then you get out to the next level of detail. I think what we shared with you before is that we have focused on larger customers and we come down market from there. The question wasn't that we are now, for the first time, adding new logos, the new logo for more apps. As we are adding more and more people, the higher end of the market is fairly well covered. You have to add more account execs, they actually naturally going to the next level of the stuff. And the next layer, 2k to 10k accounts, they actually have fairly limited coverage. It's fairly limited installed base. So by default, they end up getting more new logos and very few current customers, it's naturally going to take us in that direction. As 1 thing we told you. The second thing we said is, we are looking at more focused incentives for new logos here. In the past, we have done some here looking at doing more because it's a good opportunity for us, Mike.

Well, I mean, first off, that's just a space where there's most new logos exist. So it's a lot of fertile hunting ground. But every enterprise company goes through this where you sell you add customers and then it's a lot easier to do upsells, especially when you have a platform like we have.

So everybody wants to work on existing customers because it's just easier to get deals done. It's just -- it's easier, right? And so I think historically, we've been too lenient on how we set up the territories, and that space between the 2,000 and 10,000 has just gotten ignored.

And then we kind of realized we brought some outside help to help us analyze the numbers. And we said, wow, this is a big opportunity. We've got to go back to this and actually focus the territories. Because when you do territory planning, if you set up the territory so that there's no way you can make your number unless you sell these new logos, you get that energy and everybody starts rowing the boat and not order to go get that.

And then the alignment with marketing and how you spend money based on existing customers, I mean our previous regime was really focused on upsells. Now we're just balancing that out.

One point of just clarification. You've heard us talk about the 20,000 plus or minus, largest companies in the world. this population is included in that 20,000. So it's not like we're all of a sudden going to a completely different ideal customer profile, right? We're really talking about the same population of companies that we have been for many times, just making sure we have appropriate coverage and focus. .

Great. Do you want to close this out Jay? .

Yes. So I hope you heard that we see a massive opportunity. The market is getting hotter and hotter I mean there's nothing harder than cyber in today's world, right? And having a platform that's expanding and growing at a faster pace and the go-to-market engine focus on a consent pretty well positioned, pretty excited to really serve our customers and keep on innovating. So thank you for joining us, and look forward to working with you in coming sessions. Thank you.

Jay, thank you so much for joining our conference. I really waited for this session because I read an interview with you that you believe that investors -- you don't do a good enough job to explain your company to investors. So I want to give you the opportunity to explain your company.

And the first thing I want to start with, normally, I don't ask about the quarter, but I do want to ask about the quarter because I thought there was a difference between the way you performed financially and the way the stock reacted. What is misunderstood by investors in your company's performance?

Okay. So first of all, as you saw, the Q3 performance was very good. We beat all metrics that the investors look for. I think for going forward guidance, there're 2 aspects to it. One was Q4 and second was fiscal '27. And the 2 factors for us to be more cautious about the guidance, one was we had a couple of changes where 2 leaders reporting to the CRO moved for 2 various reasons, one for personal reason, understood, and second was an opportunity at a pre-IPO AI company. And when leaders leave, we want to make sure the transition involved and they will have some impact that factor it. And second was the Red Canary customer base. We have built a new product combining Red Canary technology with our technology. And that new solution is getting showcased next week at our annual Zscaler conference.

And the uptake of those customers, we still need to understand. So keeping those 2 things in mind, we kind of set the expectation at what my CFO will call as a prudent level. But in terms of the external market factors, the market for cyber has never been hotter. Methos has further put fuel to the fire. They're probably the biggest tailwind since COVID for our company. And this platform has gotten bigger and bigger, loyal, happy customers. And we have essentially gone through the transition of sales that started about a little over 2 years ago, and we've gone through most of it. And we look forward to expecting in Q4 and '27.

So maybe let's start with if you can articulate your target markets, meaning what are the opportunities you're going after? No, I'm not talking about the quarter thing. It could be a 3-year, 5-year, but what are the opportunities you're going after? And why are you well positioned for these opportunities?

And it also relates to the question you asked, what investors don't understand, okay? So investors understand mature established market well. A firewall is a firewall. It doesn't need to be explained much. A router is a router. Maybe the feeds and speeds are better, okay? When you bring transformation, you bring totally new changes, it takes some time to explain itself. And then what happens is the established incumbents fight back because they get disrupted. They like to say we do that, too. It's like internal combustion engine car companies fighting against electric car and say, forget it, I'm better, okay? That's the fight that goes on. So take one example. Zero Trust is fundamental, with methos it's becoming more and more important because there will be more breaches. It's given because you can't be able to patch all the way. And Zero Trust will make sure that only certain parties talk to certain parties, the breaches don't spread, the blast radius will become smaller.

Many times, people think that just because a lot of me-toos are trying to say, oh, we do Zero Trust too. And this fancy new 4-letter word SASE, we do SASE too. First of all, SASE and Zero Trust aren't the same. SASE allows SD-WAN lateral movement and the like. And then the second part is Zero Trust started with users. We made user Zero Trust Phase 1. Then we made cloud workload Zero Trust. What's my competition for Zero Trust cloud? 30-year-old firewall technology, East, West, North or traffic. Then we made branches Zero Trust. Each branch is like an island. There's no lateral movement and then device in a branch. We call it Zero Trust Everywhere story. That Zero Trust Everywhere is just beginning to take off.

We shared with you at the earnings that now we have over 700 customers doing Zero Trust Everywhere, which means users, branches, cloud workloads and devices. That same number was 550 last quarter. So big opportunities to make Zero Trust Everywhere, market #1. And there's really not a real competition from firewall vendors out there. It's only probably on the lower end of the market because the high-end customers generally are more savvy, they get it.

The second big opportunity is data security. For data security, you need to sit in line as a proxy architecture and inspect the stuff. Firewall vendors don't do data security well, they're not a proxy architecture. We have over $0.5 billion ARR business. If it were an independent company, it will be the largest data security company perhaps and still growing over 30% year-over-year. We do that stuff very well, okay? That's -- so those are one set of areas. Then look at the newer areas. Securing -- before I go there, the most important Zero Trust next area is Zero Trust for AI agents. Search for Anthropic's white paper on Zero Trust for AI agents. It just came out 2 days ago. I looked at and said, they wrote what I would have written, literally. They understand it. Their stuff was agent-to-agent communication must happen through Zero Trust, not through the firewall. We actually have been building that solution. We plan to launch it next week at our user conference, and there'll be millions and millions of user -- agents, and they need to be secured. That's an opportunity.

Then securing AI infrastructure and applications and models. We have been building that solution. We did an acquisition. And in January, we launched our integrated solution. Our customers want integrated solution not 5 different vendors. And that solution is taking off quite well. We already exceeded $100 million in bookings for that solution, a big opportunity for us. And then agentic SecOps. While there are many players in agentic SecOps, AI actually is useful for building SecOps. There are probably 500 companies, start-ups in agentic SecOps, maybe more. And because they think they're going to build it easily, the advantage will be to vendors who actually have telemetry and metadata to do it.

We are in line. We probably have the most valuable data from communication. We are on endpoints. We have endpoint telemetry. All authentication goes through it. We have identity telemetry. So our customers are saying, you got the data. I don't need to pass data to someone to build the SIEM. I want really output that can be done directly with us. So that's an opportunity for us as well. So there's no lack of products, and we do products carefully. We don't go on a buying spree for I mean to say, A, B, C. They're well integrated. They're part of the story.

What makes you be successful in these areas, meaning the current position that you have with customers, what are the parts that can be levered into the new areas?

Yes. So first of all, Zero Trust Everywhere is expanding from what we have. It's very, very natural. When customers go from Zero Trust or users to Zero Trust Everywhere, the ARR either becomes 2x or 3x, okay? Similarly -- and we are natural. If you are a user of Zero Trust, natural to the next. Data security. We're sitting in line inspecting traffic. Most of the time, data leaks through the Internet. We are the natural player to be able to do that. If you think about for agentic stuff, it's natural to be do that. And if you think about the SecOps, SecOps is driven by the fact that we got all the telemetry and metadata. We can do that, but one more point. In minutes, I can figure out some of the new threats. I can do a close feedback system to the in-line system to block those threats in near real time that doesn't happen otherwise. So very synergistic platform expansion.

Yes. Is there a risk of slow take rate? Meaning what I'm referring to is the majority of the revenues today are ZIA and ZPA. You have new areas you're going after. Is there a risk of a kind of transition period?

So if you look at the ARR or emerging growth and new ACV has been growing rapidly. And we have been giving some stats, for example, great to see data security agree so well. We have seen in the past couple of years, we used to do emerging products and some of them already emerged. So we stopped doing emerging. The emerging went from about 8% or 9% to 30% in 2, 3 years. That's a remarkable expansion thing. So if the products are synergistic and the decision-makers are similar, it becomes meaningful.

If decision-makers are very different, it becomes a lot hard. That's how we are [ selecting ] our products. One more thing as products have expanded, we now have -- do have specialty overlay salespeople who can go deeper in certain areas, but they work with account execs.

Got it. You mentioned data security, and this has been a focus of yours for quite a few years. Double-click on this market, meaning what is attractive within -- data security is a big space. What is attractive? And what are you addressing within data security?

About 5, 6 years ago, we only used to do DLP, in-line DLP. And our large customers said, doing data security with 1 vendor is hard enough, if I buy 3 products from 3 vendors, it will be impossible. Zscaler, you should focus on building a complete platform for data security. And that's when we expanded to CASB SaaS security. Then we added endpoint DLP, added e-mail DLP, and we added cloud security, S3 buckets and all. Recently, we added -- last year, we added DSPM, discovery, classification of data. So it is the most comprehensive integrated platform. We think we have a big edge. We have very good uptake by our customers. So we're very pleased with the performance.

And again, what -- the same question I asked you before. What makes you better positioned for this market given that it is being addressed by other players as well?

Not really that much. Think of which SASE vendor does data security very well, not a whole lot. There is a class of vendors coming from the start-up side, they call themselves DSPM. DSPM, one more 4-letter acronym, it's called data security posture management. At a simplistic level, it does 2 things. One, it allows you to discover data. Where is my data, data center, AWS, Snowflake, wherever. And then it helps you classify the data. That's important. But it doesn't do data DLP, okay?

You do classification and discovery, then you combine it with a DLP and then it becomes a complete solution. We came from DLP side. We added DSPM, so we have a complete solution. And generally, there's a lot of wording about DSPM is important. It is important. And you hear about some of those vendors, but they will have to do DLP to be successful in data security, and we are ahead of anyone in this area.

Got it.

And then the customer engagement tracker working with us. DLP should only be done with somebody who is already sitting in the traffic path. Some new first vendor to come and say, put me in the traffic path is a big ask.

Got it. I want to go back to ZIA, ZPA just because it's a big portion of your business. How is the growth like? And what are the drivers? I mean, when I talk to Cisco or Fortinet -- yesterday, I hosted Ken Xie and Ken said, I'm 1/3 of the price. He said, that's my thing. I'm bundling it together with my firewall. I'm writing on top of the firewall, I'm 1/3 of the price. What kind of a disruption do you see from companies who are trying to bundle firewall with SASE or SSE?

So 2 things. First, if you want product functionality A and you get B, I'm not sure in cyber, if someone gives it to me free, I won't take it. So if it doesn't work -- so they are all firewall functionality. There's nothing zero trust about it. In many product areas, good enough is good enough. Perhaps the HR system, now, they are pretty important, but...

Not as important as cyber.

Yes. I'm okay with a good enough HR system, but I'm not okay with a good enough cybersecurity. CISOs, CEOs have to think that if I get compromised, what's the consequence. That's point Number One, okay? Point Number Two is that customers are understanding more and more the need for Zero Trust. I think the wind is not towards I can give you cheap firewalls. Now it is true that lately, in my view, firewall vendors have been helped from 2 things. One, the prices have gone up. Heck, if you raise the price 15%, if you beat the quarter by 15%, what's a big deal about it, right? It's external factor that brought you in. Two is, I think they're also getting some tailwind from the AI data centers being built. But we are clearly seeing Zero Trust momentum building up. Our growth should not be looked at ZIA, ZPA for users alone. That's a starting point. It's not a core versus noncore. Our product to stay Zero Trust users to branch to cloud, how well is the overall portfolio growing is really an important part of it. And overall, we're doing quite well.

And is there -- just because of the fact that there is more competition today on SASE, even inferior solutions, but Cisco is in the market and Fortinet is in the market and CheckPoint is in the market. These are new players. They were not in this market before. Does it translate into pricing pressure or shorter duration of contracts? Or do you see any impact of the new competition?

So on the high end of the market, which we actually do extremely well, we don't for 2 reasons. One, they understand architecture value. But many times, procurement likes to bring someone in even just to put pressure on the pricing. This is how many times the dialogue goes in. A firewall vendor goes in and say, you're spending $20 million with me, Mr. Customer. Just you need to expand, here's $5 million more. And for $2 million, I'll give you what Zscaler has for free. And they're probably paying us $5 million in that account. And we are able to go in and say, oh, you're spending $20 million on firewall, which is becoming like mainframes. The future is not firewalls. So what if I bring that number down from $20 million to $10 million in 15 months or 12 months and rather than $5 million, you give me $8 million.

When the customer sees the math, it becomes a no-brainer. I had been in many of these pricing pressure calls that come from time to time say, oh, Zscaler, my budget is down 15%. I'm asking every vendor to bring the price down by 15%. That call happens as CIO from time to time. And generally, less pressure on security, but many time they come, I can say, Mr. CIO or Ms. CIO, why do you only want to reduce 15%? Why don't I help you reduce actually $5 million or $10 million. They say, really, how? Here are the things we can take out. When that discussion happens, CIO is not worried about saving $500,000. He want to save $5 million to $10 million. And we're able to open new areas and new opportunities in that area. So overall, there's not a meaningful change in pricing pressure.

Got it. I understand. Can you talk about new customers versus upsell to existing customers? What are the dynamics of gaining new customers versus the other part?

So our customers are overall very happy customers, very loyal customers. So upsell is a lot easier. In fact, quite a few times, the call comes in, the customer or CIO or CISO move from company A to B, they call us and say, "Hey, I want to bring Zscaler in." So in fact, it's probably one of the most effective lead sources for us. We got hundreds of customers who has [ CISO ] gone from company A to B to C. And early on Monday, I got an e-mail from a CISO who just joined a Fortune 10 company. And he was in a Fortune 25 company, before another one. He bought us in Company 1, Company 2. And this Company 3, he wanted to connect and he said, "Hey, I don't even have to do much work. Zscaler is already deployed here, but I can probably expand it."

So expansion is obviously easier when you have a good customer base. New logos do take more effort. There have been discussion inside the company over the past few years, do we give more attractive comp plan for new versus old? And part of the discussion was, do I want people to do too much focus on new at the cost of upsell? If you are a company with a small portfolio of products, you must get new logos to grow. If your platform keeps on growing, you have 2 ways to grow, upsell and new. We have both of those opportunities for us. If you think about the total customers, about -- over 45% of Fortune 500. And if you go to the bigger enterprise level from 2,000 up, 2,000 is a threshold for enterprises and there are about 20,000 enterprises and about 4,500 are customers. That's about 23%. This is a sizable market to go after new.

So one of the things we're changing going forward this year is as we add new salespeople, when company is growing, you need to add salespeople. So we have more being added between 2,000 to 10,000 space, which is largely new logos because our penetration there is limited. So we're adding new reps or new logo focus. We are making sales comp plan more attractive for new logos. And we're also going to be more targeted with VARs because at that end of the market, you do need to work with VARs to have proper coverage.

I asked you about competition, and I forgot to ask you about Microsoft. So I want to go back to it and see, do you feel the pressure from Microsoft? I mean they have very disruptive pricing, but product quality is not the same. So how do you see Microsoft in the market?

So over the years, Microsoft has been a great partner. We were the one who helped Office 365, local breakout big time. That's when we got the highest level relationship with Microsoft. We're integrated with Endpoint, EDR, Identity, Microsoft Sentinel. And about 3.5 years ago, they launched a competitive product. I was kidding -- not kidding really telling Microsoft, you guys even stole my name. It became Internet Access or Private Access. At that time, as I talked to Satya, Scott Guthrie, and they basically said, look, you've got a 15-year lead over us. We need to offer the product. But working with our sales team, my sales team's quota is $100 million. Security in a given customer will be $5 million. They're not focused on security. If your product is good, the customer likes it, they'll be fine. So 3.5 years later, I don't even recall last time I had to say, man, we are competing with Microsoft. Now my data points are more on the higher end of the market because that's where we spend more time. But it is not a meaningful factor.

Got it. So working on the portfolio is one challenge, and you've done great. You have a very articulate product strategy. The second part of the battle is working on go-to-market. So talk about the efforts on go-to-market, the success stories, the challenges, kind of give us the whole picture of how you evolve your sales organization to address the new opportunities?

Yes. So about 2.5 years ago, we made the decision that we need to move from opportunity-centric sales thing to account-centric stuff. We brought Mike Rich, CRO. He has done a great job in making changes. We made big changes at that time. We wanted leadership and salespeople who knew how to deal with large accounts brought in from the ServiceNow model, which was successfully done. So over the past 2 years, we have actually successfully made all the changes. I mean, if you look at the numbers we delivered over the past 2 years, they're pretty impressive. Now they slowed the absolute growth in net new ARRs, slowed down in '24. It picked up in '25. In '26, as you saw the numbers, net new ARR growth has gone up, coming -- going from '24 in the some 0, 1% range to 7%. The first half of '26 was about 10%. The last quarter was 14%. We did pretty good. Now we want to do better.

I mean I'm an ambitious person. I mean, I have big aspirations, so really need to keep on moving in that direction. We did have a little setback with a couple of sales changes that we have factored in the guidance. But the market is good. Essentially as we pass through the transition of a couple of sales leaders and changes, everything is pretty well aligned. GSIs, we made a lot of good progress. They're working, doing some very good deals with us.

Got it. What are your challenges? I have -- I want to ask you other questions, but before that, I want to understand where do you put your focus? Meaning what are the things that you think you need to address for the next 2 to 3 years, 5 years?

In general, I have only 2 focus areas: build amazing products, sell and support to customers, okay? Now in that context, if you look at -- I think we got the building product and platform, we got it pretty well under control. We have done very targeted tuck-ins acquisition, which actually add some differentiation to our platform. The Symmetry acquisition was especially very, very good. On the go-to-market side, I think the #1 thing I wonder about fixing is the FUD being created by known Zero Trust companies to confuse the market. I know why they're doing it because they need to protect themselves. But in some ways, it's kind of doing disservice to enterprises who believe that this is real security when it's not. So that's one initiative.

We have a CMO who is very good. Now, we've got a number of initiatives in marketing to create brand and awareness. That's number one. Number two, I think related to market somehow has discounted the role we are playing in AI security. And if you really think about that, the biggest thing to be done in AI security among all these things is securing agent-to-agent communication. Nobody is better positioned for that. A number of other areas, can you give me visibility into AI products? Of course, everyone will do it. It will become like CASB kind of stuff. Can you do red teaming? Everyone will do red teaming about it. And we have it in our portfolio. We have to have it. But the hardest problem to solve is communication of agents. And that's what we built.

We'll be launching it next week. And I wonder about why has market discounted us for AI. I know the one thing that comes to mind is this methos thing, while we have been part of the Glasswing project from day 1, just that when the press release came out at short notice, our name was missing, and we had to go, hey -- I mean, well, they did it on their own time in a short window, we missed out. And then the following week or so, we're able to go to them, get approval to go out there. But we missed the window and some of the investors and market felt that. Only 2 companies that are listed there are actually AI ahead in the market. I think coming quarters will prove that we'll generate some real, real numbers in AI security to convince investors that we are the real game.

Got it. Most companies in the space are offering some kind of flex programs and you have your Z-Flex. Talk about the importance of it to the customer, to you? And then what's the take rate of Z-Flex?

Z-Flex has exceeded all of our expectations. It's probably 4 quarters or less than 4 quarters, about 3, 4 quarters.

Maybe we'll start what is it? For those that don't know, what is Z-Flex?

Yes. Thank you. What's the Flex program, as name implies Z-Flex, it gives you flexibility to buy certain products and being able to even swap certain products. Otherwise, the customer says, Zscaler, you got these 8 data security products. I'm not sure if I want A or B or C. I'm going to test and test and test. It could take many quarters. Now we are able to say, look, you can select X products, and we'll give you the ability to swap and swap is linked to similar price ranges, so to speak. So they can swap. That's number one. Number two, they said, I want 6 products, but I'm not ready to roll out all 6. It will take me time. So I want to be able to do these 4 and then do 2. So we gave you staggered, the ramp product. We had done ramps before. It just formalizes ramps as they are needed.

And number three, if I want to buy a new product, there's a rate card already available. So this thing removes the procurement cycles back and forth. All that stuff makes it easy. Now our program is fairly conservative. We basically are recognizing the revenue for the next 12 months or ARR for the next 12 months. So it's -- but it's worked very well. We just got past $1 billion in the Z-Flex booking, which is good. It's good for customers. It's good for us. And the deal size, length has gone up. Most of the Z-Flex deals are 5-year deals.

Is there a risk -- and I've seen it with CrowdStrike before. Is there a risk that Z-Flex makes it very easy for customers to try and test and systems, et cetera. And upon renewal, though, you'll see a slowdown, meaning companies are trying it, companies are testing it, but maybe renewal rate won't be as strong. Like what's the risk that what we're seeing today, some of the growth we're seeing today is really customers trying new products?

So first of all, Z-Flex is a headwind for ARR recognition.

Got it.

Okay. Because some of that is staggered, okay. Number two, our focus as a company has been to make sure the products we sold are deployed. I get a personal review on a monthly basis to look for deployed versus undeployed products. Many of the compensation of product leaders are linked to deployment. Our customer support team, the deployment team is linked to it. So I think companies should be careful, but we are already very focused on making sure products we sell get deployed.

Okay. We officially ran out of time, and I didn't leave enough time for questions. But thank you very much, Jay. It has been great, and I'm always happy to host you at our conference.

Thank you. I think if I leave the last word for you, methos will probably drive Zero Trust need faster than anything else that's driven out there. And we have the right products, and we intend to do the right execution.

Absolutely. Great. Thank you.

Thank you.

Great. Good afternoon, everybody, and thanks a lot for joining us. It's a pleasure of hosting CFO of Zscaler, Kevin Rubin. Really appreciate you joining us today. To begin, just level set, we'll take questions in the room as we come. So feel free to send it or e-mail. It's going to be a fireside chat format.

So just to jump in, right, I wanted to -- just for audience not as familiar with Zscaler story or maybe still sort of know you as the original [ SaaS module ] story. What the most important, I would say, conceptual takeaway from the last quarter was increasingly becoming a control plane, not just for the users and applications, but getting more agents and workloads. So just how should investors really think about Zscaler as a platform identity?

Sure. And thank you for having us. So Zscaler started as a Zero Trust platform for users that included Internet access and then shortly thereafter, private application access and have since expanded into Cloud, so being able to provide Zero Trust for workloads and devices sitting in Branch locations. And then more recently, we're going to be announcing Zero Trust for agents. So providing the same principles that we applied for users, devices, workloads, also providing it for agent-to-agent and machine-to-machine communication.

We have our user conference next week in Vegas, Zenith Live, and you'll hear a lot more about AI, agentic and our integrated SecOps product next week. So we'll be talking more about it. But essentially, if you think about Zero Trust, provide the least permissioning to solve a particular ask or use, don't give access to somebody to jump on a corporate network and potentially have lateral movement access to do bad things. And so we've applied those principles through our Zero Trust Exchange to really provide one-to-one communication paths between users, applications, between workload and workload, between devices and devices. And so it's just a very differentiated approach to network security.

Great. Very helpful. Just on the topic of AI, there's been sort of product debate around frontier models and AI native start-ups, especially post the Mythos announcement by Anthropic. Just can you elaborate like as AI traffic sort of expands as more and more agents are coming online and even the attackers are moving at machine speed, how does that really expand the need for what you offer, which is like an in-line enforcement layer? And how does that opportunity multiply in the agentic era as you go?

Yes. So as Mythos and other similar frontier models demonstrate an ability to identify vulnerabilities at a rate and pace that far exceeds what we've seen today, we believe the best security and the best response to that is to hide your applications and limit your lateral movement. So ultimately, your blast radius can go down to one infected device and one only, doesn't have an opportunity to permeate the rest of your network.

If you look at large organizations today, they already have a backlog of patches for existing known vulnerabilities that they are incapable of handling in any reasonable period of time. So they go through a process of prioritizing which patches I'm going to apply when and attempt to do that without disrupting their organizations and their business.

What Mythos and other models similarly are identifying is a volume of vulnerabilities that already exceed what they can handle today, you're just piling on top an order of magnitude more sets of vulnerabilities and ultimate patches that would need to get run, and it's just incredibly overwhelming.

And so our approach and our answer to that is if you adopt Zscaler Zero Trust approach, you end up hiding all of your public-facing applications, so they cannot be seen. If you cannot reach the application, you cannot breach it. And so hide it and then only apply -- only allow those sessions to be activated based on least permissioning. And so those are the conversations that we are having with our customers and prospects today in response to just an unprecedented level of identification of vulnerabilities.

That's very helpful, Kevin. And I know you did highlight the customer conversations with CISOs and CIOs, how that's really shifted post-Mythos. Just in terms of big picture, right, what are you seeing as the budget trends for broader cybersecurity, right, in light of Mythos? Are you seeing CIOs starting to kind of expand and unlock more budgets because of the extra risk or additional risk that AI is driving? And are you seeing, especially to your point around the exposure risk? And recently, there was findings of almost 10,000 high critical vulnerabilities. How is all of that kind of changing the budget landscape for you?

Look, it's incredibly challenging for a CIO who has largely lived within a fixed budget domain for a long period of time. And all of a sudden, there's 2 fundamental dynamics. You've got a need for their organizations to deploy AI and to do so quickly. I don't think I've met a CEO who has told an organization to slow down and go forward with AI slowly. It's -- let's get it deployed through the organization and take advantage of it as quick as possible. And so they're reacting to a desire by their business to deploy AI.

And at the same time, they have to then understand how do I protect this AI use in a way that protects my organization. And what we've seen is that companies have rushed to be able to deploy AI technologies within their organizations, and then they're coming back behind that to actually provide security for the ultimate use.

And so we are working with customers via our AI Protect suite of products to be able to provide protection for AI use. So we offer the ability to first identify and find all of the AI assets that are being used within the organization, not just the obvious ones that you're intentionally deploying, but applications that are leveraging AI in the background to be able to do more work within those applications. So we're able to identify the landscape of threat that you may have within your organization as it relates to AI.

The second step then is to be able to actually protect the communication path going back and forth between AI. What data do you want to share, what data do you not want to share. And so we have an AI Guardrails capability that will allow you to monitor and inspect that communication in real time to be able to ensure that sensitive data is not being exposed to models when it shouldn't be and responses back from those models are business appropriate for your particular use.

And then finally, we have an AI Red Teaming solution that allows you to continually monitor those models and understand how those models may change, evolve and drift over time. Fairly new products. We announced those as a collective suite of products at the end of January of this year, but they're actually doing quite well, and there's a lot of interest around just generally being able to protect AI.

Great. Really helpful. And you mentioned about the incremental sort of AI security demand sort of in reaction to -- react to deploying AI, as you said, sense of urgency to deploy across enterprises. We've also heard from some of the channel partners, some of the customers out there that people are closely looking at sort of longer duration architectural reevaluation as well, right? So does frontier AI ultimately also - and is it getting evident in your conversation, accelerate this migration away from -- and as you guys have talked about as a lot of the Zero Trust SASE vendors talked about from perimeter-based sort of firewall appliance-based approach.

I suspect that as we kind of play this forward and organizations recognize the exposure that exists with traditional hardware approaches, network security approaches to AI and what they will have to do to be able to patch and protect their environments, it's only going to increase the validation of moving to a solution like Zscaler. We already have an incredible ROI for customers that choose to adopt Zscaler in lieu of traditional network-based security.

And as the burden and overhead of continuing to manage that, you called it perimeter base, I just used the term network security. I think it's just going to make it that much more compelling. And we have had significant inbound conversations as this Mythos, and Mythos-like exposure has been identified because companies are realizing that there's got to be a better way for us to protect our environments without constantly having to chase our tails and continually patch and chase these vulnerabilities. So we do think that it is a significant tailwind and opportunity for our business.

Great. Very helpful. And just on a high level, the impacts on your revenue model, right? I mean, we all understand like the seat model, of course, will take a sort of a backseat as we go into this agentic era. And you have talked about how, as you said, right, the AI traffic, the workloads, the agents, all those interactions are likely to accelerate materially. Like just how should we and investors think about this model holistically evolving from this human seat-based towards more sort of nonseat machine agent-based?

Yes. So we've actually been exposing some additional color into the texture of our new ACV in a given quarter, specifically to give insights into how much of that ACV is coming through traditional seat-based pricing versus more metered pricing. And it increased in this last quarter to about 30%. So we are seeing a continued increase in non-seat-based priced products in the market.

The AI and agentic technologies that I mentioned earlier are not seat-based oriented products by the nature of those products. So they are much more aligned to consumption and how much traffic is actually being exchanged. For AI, it likely is tokens, right? That will be the unit of value that gets monetized. For some of our other products, it may be just traffic and consumption. For our Branch device by way of example, that starts to look at assets and assets communicating back to other applications and company resources.

So we have continued to see -- just continued distribution of our business away from seat-based pricing. And when it comes to agent-to-agent, we think that, that traffic will be significantly greater than what we've seen with user traffic, and we expect to monetize that through tokens and consumption as well.

Great. Thanks for laying that out for us. And just I wanted to get back to the setup into fiscal '27 because as you know, the market reaction intensely seem to focus on the guide or the early look, as you said. When you think about that prudent framing, can you just help us unpack like how you're thinking about the execution piece, right, which is kind of tied to the go-to-market transition versus like there are a lot of moving pieces, as you said, around SecOps, how they're looking at the budgets, the deployment velocity, and then, of course, the agentic and AI monetization, if you can help us unpack please.

Yes. So I thought it was important in this last call that we provide an early look as to what we were seeing and giving you a perspective on growth rates into fiscal '27. We wanted to make sure that you were aligned with us in terms of how we saw our business. Against that backdrop, I provided 2 fundamental factors that were affecting how I looked at this early look into '27.

One was we are replacing 2 senior sales leaders within Mike's organization. You may recall, Mike joined us a couple 2.5 years ago. This is his second full year running our go-to-market strategy, soup to nuts. One of his leaders got an opportunity with an AI pre-IPO startup and chose to pursue it after spending, I think, almost a decade with us, so a very long-term tenured senior leader. And so he has moved on. And another leader that was a direct report to Mike is also departing the organization. And so we will be welcoming 2 new leaders into the organization. One is an internal promotion into this new role. The other will likely be an outside hire. And given that these are 2 senior individuals, I just took a cautious approach to what that would imply for fiscal '27. These are not the only directs to Mike, but he doesn't have a large set of directs as well. So just in the interest of being prudent, we took a cautious approach.

The other dynamic affecting our perspective on '27 growth is the pace of uptake of the integrated SecOps offering that you'll hear more about next week at Zenith Live. But if -- but fundamentally, we will be coming to market with an integrated SecOps solution that will leverage our existing technology from a data fabric perspective as well as our very high fidelity rich data set. And that will be the migration from the old Red Canary product into our integrated solution.

And what I don't know and what I'm again taking just a cautious view is what does the pace of that uptake look like going into next year. So those are the 2 fundamental dynamics as I looked at fiscal '27 growth. Now aside from that, we have a significant number of opportunities that could build momentum, and we're very excited about. We've talked about AI and just the opportunity that exists, whether it's a catalyst to folks choosing to adopt Zero Trust or it's actual providing security for AI, which, again, is products that we just recently put into market. So they're young, but obviously high-growing products. So -- but that's the tension and balance that we had to strike as we thought about the early look we provided.

Yes. No, clearly, as you said, like the headline early look seems to obscure some of the healthier sort of underlying signs and signals you just talked about, and especially around the health of the upsell pipeline, the AI pipeline and the nonseat business. So just -- I know historically, you guys have often guided prudently. I remember the times around the Branch Connector launch around the cloud before they showed that sort of inflection point, right? So should investors think about central monetization of AI Protect, SecOps, agentic exchange, like you have all of those stacked together in a similar way where like you have the architecture, you have the demand already there and just it's a matter of like timing and scale of monetization. Just curious, how would you...

I mean, I think the answer is a little bit different depending on the suite of products. AI is very dynamic. It's changing frequently, as you know. We are very confident in how we're approaching AI, both in terms of securing it as well as partnering with the large frontier model companies. We are part of Glasswing and DayBreak. And so that does give us a good perspective into what is coming and how we should think about releases like that.

So we're very bullish about the AI opportunity, but it does take time to build products in market are fairly early. I think as we think about the SecOps product, to me, it's really a pace of uptake question, right? We know we have customers that are on the existing product. We know that there's an incredible amount of value and insight we can provide. So how that rolls out and what that uptake looks like to me is kind of the measure of success going forward.

And then not to mention Cloud and Branch. I mean those are 2 very well-performing products that kind of build out the suite of Zero Trust Everywhere, right, moving from users to those 2 particular products. And then soon, we'll have agents. So I think there's a lot of opportunity for us to see building momentum going into next year.

So interesting, you mentioned about the Project Glasswing partnership and the potential for commercialization and monetization. A couple of partners, some of your larger peers have talked about some of the frameworks around that. Can you just kind of at a very high level, sort of lay it out like how do you plan to utilize and leverage this partnership to kind of drive the commercialization? And what does that look like as a monetization model?

Yes. I mean it's a little premature for me to do that here. You'll hear a little bit more from us at Zenith Live next week. But we have had access to these models. We are part of these programs. We've been able to run them against our environments and use those as well as have meaningful conversations with customers as to how these models may affect them and where there's opportunities. So you'll hear more from us going forward.

Got it. And just to double-click a little bit on the agentic exchange, which is the reason I brought up is you and Jay have sounded and for the right reasons, as the biggest long-term sort of opportunity, right? And I'm pretty sure we're going to hear more about that and seems like. But just as a quick preview, right? I mean, if there is a kind of traffic-based monetization opportunity there, you talked about tokens, you talked about sort of transaction processing layer. It seems like the AI traffic has already kind of exploded and as you see like a lot of numbers and metrics, how should we think about like monetizing that, right? I know we are -- or you are kind of still figuring out and trying to sell in a more stable model, more durable model, but just kind of any sense of how to think about that monetization?

Yes. I mean just to level set, the agentic exchange is not actually in market yet. The principles of how we will apply it are very similar to what you've seen us do with other Zero Trust approaches. And again, you will hear more. But the concept is simple. We have over 50 million users today that we apply Zero Trust to through the Zero Trust Exchange. I would expect that number of agents and machine-to-machine connectivity is going to be orders of magnitude greater than that.

And so the ability to be able to broker those communications, inspect that traffic and provide protection is going to be very similar to what we've been able to do with the other forms of communication. And it just -- it represents an opportunity very likely greater than we've seen thus far with users branches and devices. So very excited about the potential, and I feel like we've got a scaled track record of providing the Zero Trust approach to other forms of communication.

Very helpful. I know we'll definitely wait until Zenith Live for understanding how that agentic traffic translates into kind of more monetizable metrics. But before that, like AI Protect, as you started off with, already is achieving scale, right? I mean it's already at $100 million bookings, still relatively early compared to all your other opportunities. Just wanted to understand, I mean, you gave us a good kind of sense of like what's resonating. But like across AI Protect, there are a bunch of different components and you have Guardrails, and you have ways to protect, like the prompts and which is starting to become more significant? Or are there multiple levers right now already showing growth just from perspective of like AI Protect?

Yes. So the Guardrail product has been in market longest. I think we're actually probably just coming up on a year that, that has been in market. The Asset Management and the Red Teaming are newer. Red Teaming was part of the SPLX acquisition that we competed -- completed a couple of quarters ago. And the AI Asset Management, we rolled out, I believe, in January as part of that announcement. So each of those components have had a different shelf life so far. But collectively, we're very excited about the proposition of all of them and think that they do form the basis for what companies need to be able to protect their AI use.

Just a quick follow-up. Is AI Protect eventually and even like early signs, are you seeing that becoming a stronger landing motion, because I know you are focusing on new logos. You have been talking about some of the adjustments you made in terms of prioritization of go-to-market. Is that going to be one of your big focus areas? And I believe you mentioned you're already seeing some traction there.

Yes. So the AI suite of products do not require the Zero Trust Exchange or ZIA, ZPA specifically. So it is an opportunity for us to land and have conversation with customers, and we are having traction in those conversations. Some of the other products are more tightly embedded with ZIA, ZPA. So it would be natural that, that would be the landing point for those products. But the AI products do give us an alternative kind of side door, if you will, into introducing Zscaler to prospects.

Got it. And you did announce an acquisition, Symmetry, very recently. It seems to be very much strategically aligned with your agentic exchange sort of framework. Of course, the way you described that enabling agents kind of imposing policy around it, and it seems it's an innovative way of orchestrating the context and identity. Can you just outline like how that's differentiated from a lot of these agentic identity, let's say, methods and sort of paradigms out there?

I'll answer as the finance guy. It's a fairly technical question. So Symmetry has built an access graph that allows us to understand and infer permissions that may have been inherited to an agent through its need to do work. And so it just provides us yet another differentiated way to be able to understand identity and understand permissioning so that as we are providing in-line policy enforcement, we can ensure that the agents are only accessing permissible assets and/or applications. So it's just -- it's another piece of the agentic exchange puzzle.

Got it. And I'm presuming we'll hear more like the monetization model and like that strategy at Zenith. Okay. Just in terms of Z-Flex, right, and this was one of the, I would say, marquee quarters for you guys. You closed $480 million worth of TCV this quarter, and that was like one of the big, I would say, inflection I've seen over the last year. So I know you've understood that it's about flexibility and broader platform adoption.

As other kind of larger players out there, right, I mean, I'm referring have been leveraging this strategy. Can you just talk a little bit about like what has now resonated this quarter, especially, which is driving the sort of big kind of uptake of very, very large deals? And anything particular that stands out in terms of underlying driver?

Yes. So Z-Flex is our approach to providing customers that are looking to make large commitments over a long period of time to us. The flexibility to not have to identify each piece of technology and specific counts of everything upfront today, right? So if somebody knows that they are a long-standing Zscaler customer or intend to use our technology over a long period of time, it allows them to have flexibility for the different capabilities that they can deploy over a period of time, if they want to be able to swap into other pieces of technology or flex into other pieces of technology, it provides the vehicle to do that. It gives them fixed pricing across all of our portfolio of products.

In most cases, some are -- in some cases, we only give them a menu of lesser product. But by and large, it gives them fixed pricing that they've already prenegotiated. So they don't have to go through another procurement cycle 1.5 years into a contract to procure more. It's already been baked into that contract and allows them to make those commitments with the understanding that they've got protection if they want to use different combinations of product or they want to be able to try something else that maybe they didn't anticipate at the time they made the commitment.

So it's a very flexible way for customers to consume Zscaler. And for us, it gives us certainty. It gives us that commitment. And so we're happy to provide that flexibility as well in that relationship. And we did cross $1 billion in booking this last quarter as a result of the strong Z-Flex bookings in the quarter.

I had a quick follow-up. I think in the interest of time, there's a question which came up. So you do and you mentioned about your M&A strategy as well. You have over, let's say, $3.5 billion cash in the platform, it's definitely mature and as you said, like across the board. How do you think about the capital allocation right here? I mean, I know you've been doing a bunch of like tuck-in M&As. Mostly, it's still organic execution. Are there categories where like a larger sort of move could make sense? I mean a lot of your peers have done very big, at least compared to their standards. Is there something that you have in mind or looking at it closely?

So we have really focused on teams and technology that we feel are highly complementary to the Zscaler platform that we have today. We have not sought to expand into adjacencies or by populations of customers. We've really been very focused on how do we bring complementary technologies to market faster. And that's how I would expect us to continue to operate at least in the near to midterm.

Great. That's all the time we had. Really appreciate you joining us, Kevin, and thanks a lot for being in the room. Thank you.

Thank you.

Thank you for standing by, and welcome to Zscaler's Third Quarter 2026 Earnings Conference Call. [Operator Instructions] I would now like to hand the call over to Kim Watkins, SVP of Investor Relations. Please, go ahead.

Good afternoon, and thank you for joining us today. Welcome to Zscaler's Third Quarter Fiscal 2026 Earnings Conference Call. On the call with me today are Jay Chaudhry, Chairman and CEO; and Kevin Rubin, CFO. Please note that we posted our earnings release, shareholder letter and a supplemental financial schedule to our Investor Relations website. Unless otherwise noted, all numbers we talk about today will be on an adjusted non-GAAP basis. You will find a reconciliation of GAAP to the non-GAAP financial measures in our earnings release. Before we get started, I'd like to remind you that today's discussion will contain forward-looking statements, including, but not limited to, the company's anticipated future revenue, annual recurring revenue, net new annual recurring revenue, operating margin, gross margin, operating profit, net other income, earnings per share and free cash flow margin, our customer response to our products, our expectations regarding AI and its impact on our business and customers and our market share and market opportunity and our objectives and outlook. These statements and other comments are not guarantees of future performance, but rather are subject to risks and uncertainty, some of which are beyond our control. These forward-looking statements apply as of today, and you should not rely on them as representing our views in the future. We undertake no obligation to update these statements after this call. For a more complete discussion of the risks and uncertainties, please see our filings with the SEC as well as in today's earnings release.

I also want to inform you that we'll be attending the following conferences this quarter: the Baird Global Consumer Technology and Services Conference on June 2, the Bank of America Global Technology Conference on June 3 and the FBN Virtual Technology Conference on June 15. And with that, I'll turn the call over to Jay.

Thanks, Kim, and thanks to everyone for joining us today. We delivered strong Q3 results. ARR grew 25% and non-GAAP operating margin hit an all-time high at 23%. AI is changing the nature of cybersecurity in real time, and Zscaler is the cybersecurity platform for the AI era. This is evident in our results and the reason we are so confident in our long-term potential. We offer the industry's only complete Zero Trust SASE solution, a singular Zero Trust platform across users, across cloud workloads and across branches. Our architecture is purpose-built to address the limitations of firewall-based SASE solutions and has several key differentiators. First, we hide applications and data behind our Zero Trust Exchange, making them invisible from the Internet and eliminating the attack surface and attacker can't breach what it can't reach. Hence, this architecture provides far superior cybersecurity protection for our customers.

Second, we eliminate lateral movement of attackers with our Zero Trust architecture. We only allow authorized users and workloads to access specific applications. This reduces the blast radius of a potential breach, providing better security to our customers. This stands in stark contrast to competitors with firewall-based SASE architecture that connect users to the corporate network. And once a malicious actor gains a foothold on the network, it can roam freely and systematically attempt to compromise critical applications or steal data. This is how most ransomware attacks happen. Finally, scale matters, and our cloud-native Zero Trust Exchange is the largest distributed in-line security platform in the world that spans across 160 public exchanges, processing more than 500 billion transactions per day. This gives us the best quality and quantity of telemetry data. Simply put, no other cybersecurity vendor has access to data sets with comparable fidelity and breadth. This high fidelity telemetry fuels our AI-powered security capabilities, continuously improving how we detect, prevent and stop threats. These differentiators are especially important at a time when organizations are aggressively deploying AI applications and models with growing interest in AI agents at scale. We expect it won't be long before millions of AI agents have access to organizations' mission-critical applications and sensitive data.

Today, users are the weakest link in cybersecurity. But soon, AI agents will be the weakest link because they operate at far greater speed and have far less oversight. Even a single compromised agent can move from discovery to data theft in minutes, inflicting catastrophic damage on enterprises. Making it even more challenging, new powerful frontier AI models like Mythos are finding security vulnerabilities in software at machine speed, significantly diminishing the effort, skill and time needed to breach enterprises. All enterprises already have thousands of known vulnerabilities that they haven't been able to patch. Frontier models are multiplying these unremediated vulnerabilities by as much as 10x and even more powerful models that are currently being developed will undoubtedly make it worse. Enterprises don't have the capacity to patch and update existing vulnerabilities, so backlogs are piling up faster than organizations can address them. To tackle this challenge, the market needs to take a different approach. We provide the 2 most important defenses against these vulnerabilities: one, hiding applications from attackers; and two, eliminating lateral movement at scale. This validates the architecture we pioneered.

Zscaler was built for this moment. We started with Zero Trust security for users, so users can safely access applications from anywhere. Then we expanded our exchange to provide Zero Trust security to branches, workloads and connected IoT/OT devices. Now we are expanding our exchange to secure AI agents. An important element of agentic security is to understand which agents, users and other identities are communicating with which models, applications and data sources. On May 21, we announced our intent to acquire Symmetry Systems, a company that solved this difficult problem. Symmetry provides an access graph that maps how identities, applications and other data sources connect across the enterprise. We are integrating its Access Graph technology with our Zero Trust Exchange. We are excited to share more about this at our Zenith Live user conference in Las Vegas next month. We are also partnering with Anthropic on Project Glasswing and with OpenAI as part of its Daybreak program, formerly known as Trusted Access for Cyber or TAC, which allows us to access frontier models to proactively harden our systems and deliver better security and resilience to our customers. Against this backdrop, investors have asked us -- where is the ideal place to guard against AI threats.

We are in the enviable position of having strong visibility and control across 3 critical vantage points for superior security, network, cloud and endpoint. This is indispensable in enforcing real-time policy decisions. It is a powerful advantage for our customers and an important differentiator for Zscaler. We are enhancing our go-to-market engine across multiple dimensions to help highlight this differentiated approach. For example, we continue to deepen our partnership with global system integrators or GSIs, who play a meaningful role in expanding the reach of the Zscaler platform. we are seeing strong growth in bookings through our GSI partners. We recently announced the launch of Project AI-Guardian, a strategic collaboration with key GSI partners, which will help our partners extend the Zero Trust architecture to AI assets, including AI agents. GSIs will be able to leverage Zscaler's AI Protect portfolio to build specialized AI discovery and risk mitigation services. We are also continuing to expand our cloud marketplace motion. For fiscal 2026 year-to-date, we transacted approximately $900 million in TCV through our cloud marketplaces, which more than doubled year-over-year. This is becoming a more important route to market as cloud marketplaces simplify procurement, align well with enterprise cloud commitments and increasingly support larger strategic engagements.

These investments are helping expand our reach and Zscaler's unique architecture and approach for safe adoption of AI is resonating. This is evident in my conversations with customers and partners and why I believe AI is a catalyst for our business. Let me illustrate our progress with a few customer examples. In a 7-figure upsell deal, a Fortune 500 financial technology company chose Zscaler to secure rapid enterprise adoption of AI with our AI Protect solution, which we introduced in January. AI Protect includes AI asset discovery, AI guardrails and continuous red teaming. Zscaler AI Protect provides this customer a single integrated way to discover and manage all AI assets, including shadow AI usage, enforcing safe access to approved apps and inspecting every prompt and response in real time to stop data leaks and attacks like prompt injection. This customer faced the complex challenge of securing both employee interactions with public AI apps and their own suite of custom-built AI solutions. With our AI Red Teaming and AI Guard capabilities, the customer moved from a manual reactive effort to an automated proactive approach to harden the growing number of AI applications.

For customers building their own AI models and applications, our AI red teaming solution performs continuous security assessment. Our unified user interface and deep integration of multiple products is a key differentiator. Our AI Protect solution is resonating with customers with bookings crossing $100 million over the past 12 months. We are seeing inbound requests from across our customer base, and our pipeline is robust and growing. In another customer example, we closed a 7-figure upsell with a federal agency that previously migrated from a legacy VPN architecture last year to Zscaler's Zero Trust platform. This agency is deploying Zscaler to modernize and unify its data security strategy, gaining broad coverage without the overhead of managing additional endpoint agents or the operational complexity of stitching together various data security products. With this expansion, the customer is now using 6 of Zscaler's 8 data security modules across data classification, e-mail DLP, endpoint DLP and in-line DLP, along with our GenAI security solution. The expansion underscores our broader momentum in data security, which crossed $500 million ARR, up over 30% year-over-year. As AI adoption accelerates and sensitive data increasingly resides across multiple locations, customers are reducing cost and complexity by consolidating onto our data security solution.

Turning to another customer example. During Q3, we signed the largest branch deal in Zscaler history, an 8-figure upsell with a leading health care system to deploy our unified Zero Trust Branch solution across 2,000 sites. Zero Trust Branch disrupts branch firewalls, software-defined wide area networks or SD-WAN and MPLS networks. With this win, we are displacing both a major firewall incumbent and a legacy VPN incumbent. With Zscaler, the customer is eliminating lateral threat movement in their health clinics at roughly half the cost of its prior legacy solution. We are seeing particular momentum with Zero Trust Branch, where ARR has approximately tripled year-over-year. The next customer I'll highlight is a 7-figure new logo win with a leading health care technology company for a platform-wide adoption. This deal illustrates why customers choose Zscaler over incumbent firewall vendors, the stickiness of our Zero Trust approach with CIOs and CISOs and our ability to convert a limited initial request into a comprehensive platform win. We received an inbound request for this particular customer after a senior technology leader joined from another customer, where he had a great experience deploying Zscaler.

He fully understood the difference between Zero Trust SASE and firewall-based SASE, while the initial discussion started around securing users, the company's top priority quickly became securing cloud workloads. The deal quickly grew into a comprehensive platform win, including Zero Trust Cloud, Zero Trust Branch and 4 data security modules. The last customer win I'll highlight is a large automotive manufacturer that adopted our Zero Trust Cloud solution in a 7-figure upsell deal. This is a long-time Zscaler customer whose ARR is up tenfold in the last 7 years. This quarter, the customer extended its existing Zscaler Zero Trust SASE footprint by expanding its deployment of Zero Trust Cloud, improving its security posture and securing its massive multi-cloud environment. The customer can now inspect encrypted traffic and enforce granular security policies across hundreds of previously ungoverned cloud workloads. The deal also highlights the benefits of our Zero Trust Cloud solution, which was configured in under 10 minutes during the customer's proof of concept. Zero Trust Cloud eliminates virtual firewalls in data centers and cloud environments, reducing cost and operational complexity. The strength we are seeing in Zero Trust Cloud and Zero Trust Branch is driving the growth of our Zero Trust Everywhere enterprises that purchase each of our Zero Trust Users, Zero Trust Branch and Zero Trust Cloud. We exited Q3 with more than 700 Zero Trust Everywhere enterprises versus over 550 in Q2.

Customers are recognizing that it is no longer enough to just secure their users, and our platform is the industry's only complete Zero Trust SASE solution across users, across cloud workloads and across branches. In summary, we are confident Zscaler is the cybersecurity platform for the AI era. We expect AI and Mythos-like frontier models to be one of the strongest tailwinds our business has ever seen. Our Zero Trust SASE solution enables us to hide applications and make them invisible to attackers while also eliminating lateral movement. These attributes, along with our scale, are true competitive differentiators for Zscaler. With frontier models uncovering vulnerabilities at unfathomable speeds and AI agents becoming the weakest link in cybersecurity. These differentiators have never been more important than they are today. Our approach is resonating and helping to drive significant wins that demonstrate our ability to attract new customers and further penetrate our installed base of more than 9,400 customers. Among those, we serve just 4,500 enterprises out of a potential 20,000 enterprises in our primary target market. We are confident that our innovative approach to staying ahead of threat actors will help to drive further share gains. With the significant long-term growth potential, we are well positioned to continue creating significant value for shareholders.

Now I'll hand it over to Kevin to walk through the financials.

Thanks, Jay. We delivered strong Q3 '26 results, growing revenue 25% while investing with discipline. Year-to-date, with 26% revenue growth and a 29% free cash flow margin, we achieved Rule of 55 performance. Our Q3 '26 net new ARR was $166 million, up 24%, bringing total ARR to $3.5 billion, up 25% year-over-year. Net new ARR benefited from strength in the public sector vertical, which includes state, local and federal government and health care, including an approximate 8-digit upsell at a federal agency. Net new ARR also benefited from strength of large deals in APJ, where the deal value from $1 million-plus deals increased more than 150% year-over-year. Excluding the contribution from our acquisition of Red Canary, net new ARR was $153 million, up 14% year-over-year, and total ARR was also up 21%. Red Canary exited Q3 with $127 million of ARR. We have steadily expanded our Zero Trust platform beyond users to protect branches, workloads, AI applications and now AI agents. We believe AI agents will drive a meaningful increase in machine-to-machine and agent-to-agent interactions over time. In Q3, our non-seat-based metered usage solutions delivered just over 30% of new ACV and the ARR tied to those offerings grew more than 100% year-over-year. Revenue of $850 million grew 25% year-over-year and 4% sequentially, exceeding the high end of our guidance.

We closed Q3 with 748 customers generating more than $1 million of ARR and 4,003 customers exceeding $100,000 of ARR, growing 18% and 19% year-over-year, respectively. We also set a record $1 million-plus new ACV deals for Q3. On a geographic basis, we saw strong growth from the Americas, which accounted for 56% of revenue, up approximately 31% year-over-year. EMEA accounted for 28% of revenue, up approximately 16% and APJ for 16%, up approximately 23%. Remaining performance obligation, or RPO, of approximately $6.5 billion grew approximately 30%, including approximately 46% classified as current RPO. Our go-to-market strategy is a key growth lever, enabling us to deepen customer relationships, accelerate platform adoption and expand multiyear engagements. Building on Jay's earlier comments on enhancements to our go-to-market engine, we are continuing to strengthen our position as a long-term strategic partner and driving deeper customer adoption over time through our account-centric sales motion. We saw strong momentum this quarter with Z-Flex. Z-Flex gives customers with multiyear commitments the flexibility to activate or swap modules without starting a new procurement cycle, along with premium deployment assistance and support.

This program is driving meaningful upsell, shorter sales cycles and greater forward visibility. In Q3, Z-Flex generated just over $480 million in TCV, up more than 60% quarter-over-quarter. We have delivered over $1 billion in Z-Flex TCV over the last 12 months and an average 4-year term, underscoring customers' long-term commitment to Zscaler. To share a couple of customer examples, in a 5-year 8-figure Z-Flex deal, a Fortune 500 finance and insurance customer that spends more than $5 million with us annually increased their ARR by nearly 50%, expanding module adoption across 4 existing modules and adopting 6 new modules, including our AI Protect solution. In another example, an existing 7-figure ARR Global 2000 semiconductor manufacturing customer increased their annual spend with us by 60% in a 3-year 8-figure Z-Flex deal. This customer expanded adoption across 6 existing modules and adopted 6 new modules, including our AI Protect and Zero Trust Branch solutions. Turning to operating performance. Non-GAAP gross margin was 80.7% compared to 80.3% a year ago. Non-GAAP operating income of $196 million grew $49 million or 34% as compared to $147 million last year. Non-GAAP operating margin of 23% increased 140 basis points year-over-year, demonstrating leverage on sales and marketing. Turning to the balance sheet. We ended the quarter with $3.5 billion in cash, cash equivalents and short-term investments and $1.7 billion of debt.

In Q3, we generated $198 million in operating cash flow, and CapEx was $42 million or 5% of revenue. This equates to a free cash flow margin of 16% this quarter, down from 18% last year, reflecting the timing of cash collections and a free cash flow margin of 29% year-to-date. Looking ahead, I'd like to spend a minute and provide an update on increasing memory storage and processor prices and availability. As a reminder, we purchased equipment for our data center and Zero Trust Branch appliances. To mitigate costs, we put through a price increase on our branch appliance earlier this calendar year, which we expect to flow through in the next several months. We are also being opportunistic in taking advantage of delivery of data center equipment where we can get it to lock in today's prices ahead of potential increases in the future. This is pulling forward some of the investments we expected to make in fiscal '27 into Q4. As a result, we expect higher CapEx in Q4, taking fiscal '26 CapEx to the high single digits as a percentage of revenue, up from our prior expectation of mid-single digits. Looking ahead to fiscal '27, based on higher prices we see in the market today, we expect CapEx as a percentage of revenue to increase up to 200 basis points compared to fiscal '26 levels.

We'll continue to monitor our costs and share regular updates about the impact. Turning to guidance. At the end of the third quarter, 2 sales leaders departed the company. We already appointed a replacement for one of these leaders, and we are in the late stages of hiring a leader for the other role. However, we are taking a prudent approach to our guidance during this transition. Let me provide our outlook for Q4 and full year fiscal '26. As a reminder, these numbers are all on a non-GAAP basis.

For the fourth quarter, we expect revenue of $875 million to $878 million, reflecting approximately 22% year-over-year growth; gross margin of approximately 80%; operating profit of $206 million to $208 million, up approximately 30% to 31% year-over-year; net other income of approximately $24.5 million and earnings per share of approximately $1.08 to $1.09 per share, assuming a 21% tax rate and 168 million fully diluted shares. For the full year fiscal 2026, we expect ARR of $3.740 billion to $3.749 billion or year-over-year growth of approximately 24%. This guidance implies net new ARR growth, excluding Red Canary, of approximately 9.5%. For Red Canary, we expect ARR of approximately $137 million in fiscal '26, up from our prior guidance of $130 million with net new ARR of approximately $10 million in Q4.

This includes all the business expected in each period, including fiscal '26 renewals, upsells and new logos. Revenue of $3.3295 billion to $3.3325 billion, reflecting year-over-year growth of 24.6% to 24.7%. We expect Red Canary revenue of approximately $137 million in fiscal '26, up from our prior guidance of $125 million. Operating profit of $755 million to $757 million, up approximately 30% year-over-year, up from our prior guidance of $742 million to $748 million. Earnings per share of $4.10 to $4.11, assuming a 21% tax rate and approximately 168 million fully diluted shares and free cash flow margin of approximately 22.8% to 23.3%, down from our prior expectations of 26.5% to 27%, reflecting CapEx in the high single digits as a percentage of revenue. Looking to fiscal 2027, I'd like to provide some early perspectives to better align expectations heading into our second year with ARR as our primary growth metric and following the acquisition of Red Canary. Sitting here today, our view is for total ARR and revenue growth for fiscal '27 of 16% to 17%. Looking ahead, we are excited by the opportunities we see to continue scaling our rapidly expanding AI security portfolio, accelerating Zero Trust Everywhere adoption and growing our data security revenue.

In summary, we are pleased with the results we delivered year-to-date in fiscal '26. We achieved 25% year-over-year ARR growth and record operating income. We also saw continued momentum with Z-Flex and closed a record number of $1 million-plus ARR deals for Q3. The opportunity ahead of us is substantial, and we are confident in our ability to continue driving profitable growth across multiple vectors, including product innovation, go-to-market and customer expansion and creating value for our shareholders. I want to thank our employees, customers and partners for their continued support.

With that, operator, you may now open the call for questions. Thank you.

[Operator Instructions] Our first question comes from the line of Brad Zelnick of Deutsche Bank.

A lot of strong proof points in these results. Maybe a compound question since you're limiting me to one for both Jay and Kevin. I'm just trying to better understand the sales leadership turnover that you highlighted. Jay, if you can comment at all whether these positions, the turnover was voluntary or involuntary, how senior they were and why this could have an impact when you have such a mature and resilient sales organization? And then maybe for Kevin, on the same topic, is the pipeline there and you're just assuming a lower close rate? What would your guidance have been if these leaders were still in place?

So regarding the sales leadership changes, these 2 leaders are part of our CRO, Mike Rich's team. And it is true that Mike has built a strong bench. He has built a strong sales engine. We just want to be prudent that as these changes are made, it could have impact in the short term. And that's what we are keeping in mind.

Kevin?

Yes. Thanks, Jay, and thanks, Brad. Yes, I don't have much more to offer other than we are taking a prudent approach. We do recognize when leaders of this nature change that it can have some disruptive nature to those organizations. So I'm just taking a prudent approach to how we think about those changes.

And the only other thing I'll add, Brad, is that there will be changes in leadership from time to time. And regarding these 2, we have already appointed an internal replacement for one of these. And the second one, we are making progress, and we expect to close that in the near future as well.

Our next question comes from the line of Saket Kalia of Barclays.

Kevin, maybe for you. Can we just speak to how big the 8-figure federal deal was here in Q3? I guess the question is, if we exclude that deal, how do we maybe feel about the underlying flow business, right, whether it's new or renewals, again, excluding that large deal, which was great to see. I'm just curious how you look at the business if we exclude it, if we could size that.

Yes. Thanks, Saket. I mean, look, at the highest levels, we delivered a strong Q3, and I think we're very pleased with the results of the business. As it relates to the 8-figure upsell that I called out, keep in mind that these contract wins, as you see them get reported, first of all, it is TCV. And second, that part of the -- along with the upsell was also a large part of that deal was renewals. So that's already preexisting in ARR as you think about that particular deal that you were querying.

And Saket, if I may add, you've seen in the past, if deals are unusually large, we point them out. So nothing unusual about this.

Our next question comes from the line of Joshua Tilton of Wolfe Research.

I appreciate the early look for the numbers next year. When you look at the growth and kind of what it implies for net new ARR, you guys are kind of calling flat on an organic basis from the guide for this year. Can you kind of help us maybe think about the moving pieces between where that's coming from, whether that's the traditional Zscaler business or Red Canary or just any color to help us think about the trajectory of organic net new ARR next year in that guide?

Yes, I'll go ahead and start, and Jay can append. This is how I thought about the different components as we were giving you guys an early look. And I just want to call out, we don't typically provide this type of guidance this early in the year going into the following, but we thought it was important to give you an understanding of what we're seeing. So first of all, we have a pretty strong track record of upsells, and I expect that, that is going to continue into '27. You may recall that we shared we had a net retention of 115% in Q1, and that has been fairly stable the last several quarters. The area that we haven't been performing as well as we'd like is new logo. It certainly is a large priority for us, but I did take a tempered view of new logos going into '27. And then lastly, as we think about Red Canary and its contributions, we will be rolling out the integrated SecOps solution that will be available, we would expect in '27.

What I don't know is the pace of uptake amongst the existing customers for that. So as a result, as we think about Red Canary, we are expecting Red Canary's net new ARR to grow at a slower rate than the overall business in '27 as you think about modeling. And also keep in mind that Red Canary will be included in our results going into next year as it will be fully baked into '26. So we will not be providing separate disclosure of Red Canary in '27.

If I may add, look, we already covered the fact that sales leadership position has been part of our thinking to make sure you understand that it may have some impact. But if you look at the overall market opportunity, Mythos. has shed some more light on it. Every CIO, every CISO I talk to is talking about protecting against Mythos. And interesting part is while Mythos is about finding new software vulnerabilities and fixing them. People already have -- enterprise already have tons of vulnerabilities. The #1 protection they're looking at doing is hiding their attack surface. #2, they're trying to do is Zero Trust access. And we fairly uniquely provide those things. So while we were prudent in our forecast, the need for Zscaler today, especially for [indiscernible] Mythos is probably far greater than it has ever been.

Our next question comes from the line of Gregg Moskowitz of Mizuho.

You did very well this quarter in the Americas and APJ, although growth in Europe slowed. What are you guys seeing in Europe? And then just wondering if you had any commentary on the competitive environment, both in the Americas and in Europe.

So if you think of the overall competitive environment, I don't think things are very different in Europe than the rest of the world. You have seen us from time to time, Geo A may do better in a given quarter and Geo B may do better. But I think we know some of the areas of execution we need to improve. We are focused on it. And I'm pretty sure that we'll turn it around and make EMEA again a high-growth area.

Our next question comes from the line of Gabriela Borges of Goldman Sachs.

Kevin, you mentioned something interesting there when you were talking about the outlook for next year, which is that new logo growth may be tempered. And so either for yourself or for Jay, maybe just give us a sense when you dig beneath what's going on with the sales relationships with the customer pipeline, why do you think new logo growth is tempered?

Yes. Thanks, Gabriela. Just to clarify, I was not trying to suggest that new logo growth is tempering. What I was intending to convey is that my expectations relative to the early look for next year took a tempered approach to how we thought about the contributions of new logos. New logos is a strategic focus of ours. And as we go into '27, it will continue to be so. But again, I'm providing an early look here far earlier than we would normally do. And so as I looked at what we had in front of us in terms of the different components of the business, that was one area where I think we can do better.

If I may add, there's a sizable new logo opportunity for us. We have 4,500 enterprises as our customers, and those enterprises are customers with over [ 2,000 ] user seats. And out of the [ 20,000 $4,500 ] customers, that's what 23%. That means a sizable market for us to go after.

Our next question comes from the line of Brian Essex of JPMorgan.

Jay, if I could maybe ask you to peel back the layers a little bit on the impact of Mythos and OpenAI models. What we're hearing from maybe some of the channel providers is that the C-suite executives at enterprises are maybe to use their words, freaking out over the emergence of these models and the realization of how robust they are. So when you comment about some of the tailwinds in your business on the back of the emergence of these foundation models.%.

How do you see that materializing in your pipeline? It sounds like the consultants are very busy right now. But given the typical sales cycles that you have and maybe the architectural decisions that are involved with adopting your Zero Trust platform, when might we expect to see some kind of impact to pipeline revenue, billings and so forth? And how is this materializing with your conversations and how your pipeline might be materializing?

It's a very important question, Brian. We've never seen so many inbound calls come in so quickly. I got so many calls that I decided to say we will reach out to them and figure out a programmatic way of engaging with them. I don't need to dig into the detail of the concern. We all know they're very concerned, and they're looking for help. So it is interesting that most of the customers and vendors are taking the approach that let me help you find more and more vulnerabilities and try to patch that. While patching is a reasonable approach, we do not think the primary focus needs to be patching because you will never be done patching. So our recommendations are very clean, hiding your applications that I talked about, Zero Trust access, those are fundamental things that need to be done, and that's what we are helping our customers with. Now we also know that this is an area where we need to help our customers and not do ambulance chasing, okay? So we're not rushing to create opportunities. We are actually engaging with our customers in a consultative fashion to help them get out of the tough situations so they can really keep the Board and CEOs apprised of what's going on.

Now obviously, there are areas these customers need to worry about. We aren't really factoring in any meaningful impact of the new opportunities for Q4. But I do believe it will have impact in fiscal '27. The short-term impact, if I were to give you a couple of specifics, number one, many customers have ZIA for all users, but Zscaler Private Access is only for a subset of users. Now they want Zero Trust access for everyone from every location, even from the headquarters. So it's increasing interest for ZPA upsell. We have very cool deception technology. It's a decoy technology. Customers want this technology because they know that with all this happening, there will be more breaches and they want to catch that red-handed using deception like a stuff. So that's why I said sometimes I feel like it's a [ 4.0-like ] moment. Zero Trust has never been more important before. Thank you.

Our next question comes from the line of Meta Marshall of Morgan Stanley.

Great. Maybe building on a couple of the questions. Just as we think about kind of signposts that you guys are looking at for improvements on the new logo side, will that come from some of the GSI channel outreach that you guys are -- expansion that you guys are doing? Will that come with some of the new leadership, flexible pricing? Just trying to get a sense of what you think is kind of the most incremental to improve the measures that you laid out to improve the new logos.

Yes. Meta, thank you. So we have very specific plans we are formulating and they'll be an important part of our fiscal '27 plan. Number one, we have a limited coverage on the lower end of the market. When I say lower end of the enterprise market, it's generally between 2,000 and 10,000 users. As you know, the high end, our coverage is pretty strong. So adding more salespeople in that part of the market where coverage is less, number one.

Number two, the channel, especially the VAR channel plays an important role in the low end of the market. So we are creating specific programs and incentives for new logo in that area. Number three, the GSIs have been good partners. GSIs are generally good partners for large enterprises. So we are teaming with them for this area. You heard about AI-Guardian program we launched with them. That's a natural extension of the work we are doing. Number four, we have a program for major accounts. We're going to make sure we do some more focus on those teams to get new logos in addition to working on upsell. I think we feel pretty good about the opportunities to engage and do this stuff and a little bit more incentive to new logo versus we haven't done a whole lot in the past. This will probably make a difference.

0

Our next question comes from the line of Ittai Kidron of Oppenheimer & Company.

Kevin, I just want to make sure I get my bearings right around the ARR and the commentary there. Taking into account your revised guidance for the year and your preliminary view into next year, it looks like your 4Q net new ARR again is decelerating quite substantially, and it's probably also in the -- going back to mid-single-digit growth only into next year. So another deceleration. I just want to get my hands around outside of the sales leadership that you've talked about, are there any other elements to take into account here with respect to this. You guys have made a lot of work over the last couple of years kind of turning around the sales force. I understand that when 2 leaders leave, clearly, some disruption can happen. I'm just wondering if that is the only element impacting the net new ARR here? Or there are other things to take into account?

Yes. Thanks for the question. I mean, look, the 2 factors that I think are important, you certainly mentioned one in terms of the 2 leaders under Mike that we commented on. It's just the reality when you have leaders that do depart that you do -- you may see some disruption. And so I'm just taking a prudent approach to that potential disruption. The other thing that I commented on is just the pace of uptake with the integrated SecOps products. So as we think about how that rolls out and what that pace of uptake is, I'm also taking a prudent approach there. As it relates to Q4, so the guide implies 9.5% net new ARR growth on an organic basis, excluding Red Canary. And just keep in mind, that's still an acceleration over what we put up last year. So keeping everything in perspective.

Our next question comes from the line of Erik Suppiger of B. Riley Securities.

On that outlook for fiscal '27, I think you had commented last quarter that the ZIA, ZPA core products were growing in the mid-teens. What growth assumption are you assuming as you look out to '27 on those core products?

Yes. Thanks for the question. I mean it's a little bit early, and that's a fairly granular element. What I can say is that we've continued to see consistent performance this year within the ZIA, ZPA product lines, but we didn't provide that level of granularity into next year.

If I may add, quite often questions get asked about core products versus known core products as if they're 2 separate buckets altogether. I think as we have been saying, our customers are asking for Zero Trust everywhere. We started with Zero Trust for users, which is ZIA, ZPA [indiscernible]. That was part one. Now they're all moving towards Zero Trust for cloud workloads and then Zero Trust branches and IoT/OT devices. So our #1 push is to work with customers to do Zero Trust everywhere. That's a big differentiation for us and an opportunity to sell bigger deals, while others who claim to say we do what Zscaler does are barely trying to build Zero Trust users, which we matured over the last many, many years.

Our next question comes from the line of Adam Borg of Stifel.

Maybe just on Symmetry Systems, I would love to learn a little bit more about what that brings to you that you couldn't do previously around securing AI agents. And given their Access Graph technology, how are you thinking about the move into identity security more broadly? And maybe just as a quick follow-up, any color on Symmetry even SquareX in terms of contribution to revenue and ARR?

So first of all, Symmetry has built very innovative technology. So what's the problem statement? It's not really basic identity. We believe the identity of agents will really come from companies such like hyperscalers or large software companies who are actually providing platform to build agents, they are the natural place to build identity. So trying to be an identity company for agents for someone outside these big guys will be a hard thing.

We have always taken the approach of being a [indiscernible] where we will take identity from different parties. So then what Symmetry? Symmetry pioneered identity mapping to data sources. So think of this way. In a large enterprise, all these identities, maybe users, workloads, maybe other machines, they access data sources that may be sitting out there somewhere. How do you know who is accessing what, when, where? Information sits in each application logs. Symmetry pulls that information and creates a very cool visual access graph. This was a hard problem to solve.

So this was solved by a bunch of PhDs. Once you understand who talks to who, this thing can be used to enforce policy, but agentic exchange that we're building. So it's very complementary forward-looking technology. This is the kind of stuff most companies aren't thinking about. Zscaler has always taken pride in being innovative and coming with solutions that no one else has built. So our Zero Trust Exchange got extended to Zero Trust Exchange for agents, and this becomes a very useful piece to highly differentiate our exchange and provide value for data governance and policy enforcement on our customers.

Did I answer your question?

That's really helpful. And just, Kevin, maybe any color just on top line for Symmetry and SquareX.

Yes, of course, of course. So as Jay mentioned, just to level set, Symmetry is a technology and talent acquisition. The ARR is immaterial and is in the low single digits. I think you also asked about SquareX, that was also incredibly immaterial.

Our next question comes from the line of Fatima Boolani of Citi.

Kevin, just on some of the commentary with respect to the opportunities to step on the gas pedal vis-a-vis new logo acquisition and in the context of the sales leadership transitions. Can you give us a quantified and quantification of sales productivity this year and some of your expectations as you think about the complexion of the 16% to 17% guide, again, largely tied to some of your comments earlier around building out more of a capacity presence with the lower end of the enterprise. Would love a little bit more quantitative color on some of the sales productivity/sales attrition and sales hiring quantum that you're thinking about in terms of pipeline build and conversion and productivity assumptions.

Yes. Thank you for the question. Look, for Q3, this was our sixth straight quarter of sales productivity growth even on the back of some tough compares. So I think we've done -- I think we're very pleased with the level of productivity that we've seen -- continued improvement that we've seen. I would expect going forward into fiscal '27 that we'll continue to see success in driving productivity and capacity. I just caution that it's a bit early to provide quantification of how that rolls into an early look for '27.

Our next question comes from the line of Gray Powell of BTIG.

Maybe just one on the competitive front. So just from a technology perspective, how are you staying ahead of the firewall vendors who are increasingly trying to upsell Secure Service Edge into your installed base? And when you do have displacements of legacy vendors, is there like a common driver? Is it like a particular set of features? Or is it just the overall platform and just that your technology is better? Anything you could see on that front would be helpful.

This is pretty simple. If you care about real cyber protection, you know that you need Zero Trust architecture. Firewalls create trusted network. Trusted networks enable lateral movement. So our customers understand that. They started with Zero Trust for users. Now they want Zero Trust for branch where each branch becomes an island like a cafe. Firewalls don't do that. They want Zero Trust Device. Firewalls don't do that. They want Zero Trust of workloads. Workloads have been secured traditionally by 30-year-old firewall technology, East-West, North-South, firewall rules. We do it in Zero Trust fashion. So we are on a road to take our customers to Zero Trust everywhere. That is our biggest differentiator. It's not feature A or feature B. It's an architectural difference.

Our next question comes from the line of Andrew DeGasperi of BNP Paribas.

I wanted to ask about the comments on the CapEx guide. And really, what I wanted to touch on is, given you've put together price increase earlier this year, and now these costs have gone even higher, are you considering potentially doing the same maybe at the same time next year?

Yes. Thanks for the question. I'll start. Look, we are constantly looking at the balance between pricing, margin and market prices and feel pretty good at where we are. We did push through a price increase earlier this year and have not seen any implications of that. I think the world is recognizing that hardware costs have gone up. But we do periodically review pricing and where we have opportunity to increase, we do take advantage of that. I don't know that I would want to sit here today and say at this time next year, there will be another price change. It is something we look at more dynamically than that.

But this situation you're talking about is fairly unique. It's created by all these AI data centers that's causing such a shortage of so many parts. But as Kevin said, we'll look at from time to time, but this became a special factor that all of us recognize and adjusting accordingly for it.

I would now like to turn the conference back to Jay Chaudhry for closing remarks.

Sir?

Thank you for joining us on our earnings call. We look forward to seeing you at one of the upcoming investor conferences. Thank you again.

This concludes today's conference call. Thank you for participating. You may now disconnect.

Minute early, since I have to read these disclosures, and I think you would all rather hear from Zscaler. So for important research disclosures, please see Morgan Stanley research disclosures website at morganstanley.com/research disclosures. If you have any questions, please reach out to your Morgan Stanley sales representative.

I'm Meta Marshall. I cover cybersecurity here at Morgan Stanley. I'm delighted to have Zscaler here today. Jay Chaudhry, CEO, founder, multi-hyphenate, and then Kevin Rubin, CFO.

Maybe let's just kind of circle maybe to start with before we go high level on fiscal Q2 earnings. You reported 25% ARR growth, 21% organic. You raised full year ARR targets organically, but maybe organically, it was viewed as a step-down from fiscal Q1 results. Where did you see the strength in the business in fiscal Q2? And what left you kind of encouraged for the rest of the year on both businesses?

Yes, I'm happy to start. There was a couple -- a few things with respect to the first half that I think is notable. First, we mentioned we had a record number of Q2, $1 million deals for Q2 -- for Q2, which I think just underpins the transformation that we've made in our go-to-market. So that was -- I think that was a really good proof point. We also talked about Z-Flex, which is our flexible buying vehicle for our customers. We did $290 million in TCV bookings in the quarter with Z-Flex. And since we launched the program, we have about $650 million in TCV under the Z-Flex program, which, again, if you think about what it represents, it's longer-term larger deals that our customers are making with Zscaler.

So another area, I think, that just underpins some of the strength that we're seeing in the business. And then to your point, I mean, we did see strong growth in the first half of the year. My comment would be that we did raise the net new ARR expectation for the year. And it was reasonably consistent with the performance thus far. So I think we're feeling really good as we go into the back half and what we can do.

Got it. Jay, I know you addressed this on the call, but just given how topical it is, what do you think investors are missing right now about kind of this dialogue around whether AI can replace cyber versus kind of the discussions where it had been, which is that you need cyber in order to have AI?

So for us, AI is net positive opportunity. So to deploy AI security, you need to secure AI, because unlike many other technologies, AI can be a very dangerous tool. Hackers are embracing AI faster than enterprises are. So when I talk to lots and lots of CIOs, they're all doing some kind of projects out there. And they're all nervous by rolling them out, and they really look at 2 things. One, tell me what AI use is happening in my company. So understanding of getting visibility and risk associated with that. Then as they are building applications, they want to do testing, vulnerability assessment. So red teaming type of tools are needed. When they deploy them, then they want some guardrails to make sure those applications and models aren't abused.

So that's one product set, maybe call it solution we launched on January 27, we call it AI protect, integrated or comprehensive, so they can get moving with it. And the next thing they are all excited about is AI agents. Because they think that AI agents can give them the productivity that's needed. And that creates a whole new kind of challenges. Traditionally, a user has in the weakest link from a cyber point of view. It's very much expected that AI agents will be the weakest link. Imagine agents getting hacked or hijacked in your enterprise, they have access to all kind of application services. And users move slowly, agents move on the dime. Users are thousands and agents will be millions. So they want to make sure there's a policy engine that can really allow right agent to access right application or talk to right agent. And that's our core competency. So we are very excited about it. We are talking to lots of customers. We have a growing pipeline.

Got it. What we found in our checks kind of despite some of the negative commentary around AI, is that you're finally seeing kind of broader SASE adoption that's being spurred by AI. Just beyond kind of the visibility and the agent protect, just what are you seeing as the biggest catalyst today for organizations maybe moving from kind of one piece of SASE to kind of a broader SASE adoption?

So first of all, SASE is a catch all phrase, it's kind of open buffet. You pick whatever you want to pick. We like to talk more about Zero Trust security. Because that's what's trying to fix security. The security is broken because we have been doing it by trying to secure the network by building firewalls and VPNs around. When Zscaler came on the scene, and we said forget firewalls, forget network security. You don't need to trust the network, you don't need to trust the users. Everything is untrusted, Zero Trust.

You give this much trust, so you can access application A or B or C without being on the network. The first part of Zero Trust was users. Users being able to access certain applications. The next part became workloads. Workloads are somewhat like users. They talk to Internet, they talk to each other. Very natural for us to extend or exchange for workloads.

Now what's being done in the workload areas so far? Same thing that has done for 25 years. Virtual east-west firewall, virtual northward firewall. So this is the next phase we brought. And the third phase we brought is Zero Trust for branch offices. A typical compromise happens this way. A single infected machine in a branch office gets compromised. The malware moves laterally and this mesh network, brings everything down. In the world of Zscaler you don't have lateral movement, you don't have mesh network. Every device from the branch talks through our exchange. So that policy gossip to the users to workloads to branches. And the next phase is agents. We think it's a wonderful opportunity.

Okay. You addressed this on the call as well. Zscaler continues to lead to market share gains for SASE. Our checks continue to kind of point to strong positioning. Yet we get questions about the number of competitors in the market. Have you observed any changes in win rates or kind of the competitive landscape around SASE?

In the large enterprise space, which is our primary segment, that's about 20,000 employees. If anything, I would say the competition has become less. Why do you say that? Because these large enterprise customers are CIOs, CISOs, they all know us. They understand what we bring to the table. And they also listen to some of the competitive story, where underneath its virtual firewall sitting around. So they see the difference. And they're pretty savvy.

There's less -- when you go down market, there are more competitors out there. But I can tell you that any of these new competitors coming in, having talked to hundreds of large customers in the past few months. We haven't really seen any increase in competition there. The biggest focus from CIOs is, one, I want to remain safe, give me Zero Trust and help me secure my AI initiatives. Number two, can you do it with greater ROI and remove some of the cost out of it.

We are able to do both. A firewall company won't do that, even though they try to talk about platform because the biggest spend in security still is firewalls. They have to cannibalize it. We like to cannibalize it. They don't. So we are able to go out, take some of these products out, make the case very compelling. That's what's driving our growth.

Got it. We've had kind of additional discussions over kind of the past months about seat-based models, and SASE has traditionally been a seat-based model. You guys talked about 25% plus of new bookings being non-seat-based for Q2. Just how do you think about kind of that overall value proposition to customers and kind of what that method of charging for it should be?

Yes. Traditionally, most of our business has been seat-based for ZIA and ZPA, but also some of the business has been known seat based. When we do guest WiFi, there's no fixed number of seats. It's charged based on the traffic flow. When we charge third-party contractors, suppliers and customers, that's not a fixed number of seats that's linked to traffic. So we have had those things. And then you go to the next level, take some with Zero Trust workloads. Workloads it's based on number of workloads and the amount of traffic. Take data security. We have 8 modules. A couple of those modules are based on number of users, but many of these modules aren't linked to users. When I do data discovery around terabytes of data or I classify it, it's not linked to users.

So this piece of the ARR has been growing steadily, and it got to 25%, and I thought it was a good milestone. We expect it to keep on growing. The biggest new growth factor in this area will be 2 pieces. One is AI security products, which is about red teaming to guardrail, all this stuff. That's all based on tokens essentially and then AI security exchange, or agent security. They will be all token-based.

Okay. We've spent some time talking about AI security. We've spent some time talking about Zero Trust Everywhere. Kind of that third growth pillar has been the data security everywhere. Those 3 businesses have now kind of crossed the $1 billion in combined ARR, growing faster than the overall business. Just how do you see kind of maybe some of these growth pillars over the next couple of years developing?

So Zscaler, Zero through Branch has gone through an inflection point. We are seeing very rapid growth. It eliminates traditional networking, SD-WAN and the like. Five, 7 years ago, I used to say MPLS is going to disappear. People said you're crazy. Where is MPLS today? It's by and large gone. I think SD-WAN will follow the same fate too, it's a matter of time. About a year ago, when we were just getting ready to launch Branch, I used to wonder, what percent of Zscaler customer will embrace it? And what percent will say, I need to keep my SD-WAN, I love it.

Again, I've been surprised that almost every customer I have talked to, they're all ready to make Branch like an island and, no mesh network, no SD-WAN. Big opportunity. I think 3 to 5 years, people are going to say, what does this SD-WAN you think used to be? So that's big growth.

Zero Trust cloud workloads, the only way it's being secured, the only alternative to Zscaler is old-school virtual firewalls. They are hard enough and complex enough in a data center, dealing with IP addresses and all firewall is a network device based on IP. This source IP address can talk to this destination IP address. It's a nightmare in the cloud. Since with Zscaler, this VPC can talk to this VPC. It's wonderful. This solution is growing very well.

Data security. Data security, we started many years ago. We've done DLP. Then about 5, 6 years ago, some of the CIOs told me, we get tired of managing multiple DLP policies. Even one product is hard enough, dealing with policies for multiple is very hard. Zscaler, you should really offer full data security offering. So we went all the way, not just in-line DLP, e-mail DLP, endpoint DLP, SASE, SSPM and on the cloud data security. It's the most comprehensive solution. And we are in line, we are a proxy. So hence, we can inspect before the traffic goes out. If you aren't sitting in line as a proxy, you can do that. That's why you don't hear from firewall vendors that they could do amazing data security because they aren't able to properly inspect it.

So this is a big area. I mean sitting at close to $0.5 billion in ARR. If we -- if data security was an independent business for Zscaler, it would probably be the largest data security company. So it's good. The many growth engines for us that are working.

Right. Okay. Maybe turning to another one of those Red Canary. You talked about -- can you talk about the long-term strategy for this business as it kind of becomes not Red Canary separate, but kind of part of Zscaler altogether.

Maybe I can make broader comments on strategy. Kevin, you can give some color to it. As we said, we wanted to get into AI Agentic SecOps because AI can disrupt SecOps. Our customers are telling us they're paying too much money building these data lakes. And they said, we have probably the biggest set of logs, and they get put some where they pay for them. So that's how we started to move into the SecOps space. Once we have the back end, we needed the front end. We needed agentic technology SecOps agent that could really solve this problem and do it in a faster and better fashion. And Red Canary actually has that technology even though they're MDR company, they've been doing this for 10 years. They had playbooks and they took the playbooks. They had converted those playbooks into agents in production. And we are integrating the agentic technology with our back end.

So there's one AI SecOps product that we can go and disrupt the market. So it's -- that product is coming along well. Red Canary technology is integrating very well. So we think it's a good opportunity for us because, if there's one area, AI will disrupt more than others. It's a SecOps.

And then, Kevin, I mean, just in terms of it's outperformed maybe kind of the expectations you had coming into the year. So just what are you kind of seeing from the business now having owned it for a number of quarter?

Yes. So I would describe when we acquired the business, we mentioned that current year renewals were not picked up as part of our initial recognition of the original $83 million in ARR. And that was, by and large, because MDR businesses historically have higher churn rates than we have in our business. And you have a much larger company purchasing a smaller company. And keep in mind, about half of their business is in what we would describe as the commercial segment, which is companies that are generally smaller than Zscaler support or at least intend to support directly with our go-to-market resources.

So there was a lot of uncertainty from our perspective as to how much of this business would likely renew. Now that we're 6 months later, we have seen what those renewal rates have looked like. And my commentary on the earnings call was they have been elevated. So I'm glad that we took a more conservative approach to those expectations. The guidance that we have provided and updated in the last earnings call, reflects now what we've seen from a renewal perspective and what we expect at the end of the year. So yes, it was an increase driven in part by what we've seen from a renewal perspective. I don't know that I would characterize it as exceeding expectations. I think it's part of the mechanics that we came into the year with?

Okay. I mean, Jay, you've done a number of acquisitions kind of over the last 12 months. Most recently, kind of SquareX, you kind of announced for browser security. Just how do you think about kind of where the white spaces are in the platform and just how M&A will be a portion of that strategy?

We are quite disciplined about what we want to offer, what we don't want to offer. If we want to offer or be in a given market, we want to do the best. Otherwise, we won't play into the market. And we look for markets that are synergistic to us. When we went to Zero Trust Exchange area as we added ZIA, ZPA, Branch, for example, we bought this company called Airgap Networks. It rounded out our Zero Trust device story, very synergistic. Came along extremely well. Take Zero Trust for information access, secured browser. I'll share my position as compared to the market. We have been offering Zero Trust access using remote browser, which runs in the cloud and offering this functionality.

We have a pretty sizable business doing this browser business, remote browser. Now there are some use cases for unmanaged device using my own device, my own browser, people want to access information. And in many cases, it used to be third-party suppliers or customers. You could use Zscaler remote browser to work with that. There's one piece that customers are wanting. They want device posture check from security point of view. You can't do posture check unless you put something on the device. The stand-alone browser companies would say, download my full browser and you can do the check and you go direct. A couple of years ago, we actually looked at some of them, even looking at see if you could buy this company. We didn't feel right. Customers said, "I don't buy 1 more agent." And this wasn't the agent. This is the mega agent. I mean it's a massive thing.

That's one problem. The second problem was as we looked at the stuff, all these browsers are based on Chromium, which is built by Google. Chrome has become big. This is the 1 billion lines of code and so many vulnerabilities are being fixed every week. Reminds me of windows, vulnerabilities that will happen every week. And so how are we going to keep up with that? You can't. So we said we will not go into the space. What we acquired is this company has done very innovative approach using browser extensions. We are able to check the device posture. So you can actually achieve what third-party browsers are talking about without having to ask the third parties to download one more browser. I think it's an elegant approach and then we have Zero Trust on the back end.

Okay. Got it. Kevin, maybe turning back to you, Z-Flex, you mentioned kind of upfront has been a huge area of success. It's like -- can you just walk through how Z-Flex is changing customer buying behavior and kind of how it's changing your visibility on the business?

Yes. So Z-Flex maybe just by way of background, a lot of software companies have offered flexible purchase arrangements. And for us, that manifests itself in the opportunity for customers that are willing to make larger commitments over a longer period the flexibility to swap in and out of products. We offer the ability to ramp generally in the first 6 to 12 months of the contract period. And then more recently, it's actually given us some flexibility around billing arrangements as well.

So if we wanted to provide a customer with the ability to pay a little bit less today and pay a little bit more tomorrow, we can -- we have the flexibility to do that through Z-Flex. It also facilitates the prenegotiation of pricing across the suite of products that are available to the customer in Z-Flex.

So there's a lot of advantages if they want to be able to extend into our data security products. They don't need to go through another buying and procurement process. They already know what products are. They've got a negotiated price, and then they can go ahead and deploy. For us, as I mentioned, we did about $290 million in bookings of Z-Flex this last quarter. To provide a little context of the Z-Flex business for us, they're generally 7-digit deals, and they average today about 4 years.

So they are a bit longer. And when we do go into a Z-Flex arrangement, it is generally a larger ACV commitment than they were making outside of Z-Flex. So it does drive bigger deals, it drives longer-term deals. And then as I mentioned, it facilitates upsell, right? They know exactly what's available to them at what price. So we think it's an attractive opportunity for some of our most strategic customers. And over time, many of our customers.

And are you seeing kind of that module adoption, the quicker adoption of kind of the additional modules as they move into that?

It's still a little early. We've had this in market for about 3 quarters now, but it certainly continues attract more and more interest, which tells me that customers are finding a lot of value.

Got it. Kind of moving back on to the AI security front. You have GenAI Security, you have security posture management, AI guardrails, AI red teaming, agentic exchange. Just how do you think about those separate kind of pieces in an AI portfolio versus kind of overarching portfolio.

That's a great question. In fact, generally, investors are wondering what's where new market. So we have 2 main solutions for AI. One is what we call AI Protect. We launched it on January 27 after integrating a number of products we've built and a couple of products that came through the acquisition of SPLX. So AI Protect has about 4 areas of functionality. First is giving visibility. Any external, internal AI application, we can tell you what you have, what's the risk. Two is, secure access to those applications, who can access, what can be access policies by user, by group and the like. Third is securing application you have built, how do you secure them? You start with doing the red teaming for vulnerability assessments. You know what the risks are. Then our guardrail product, which inspects prompts to make sure people can do bad things like prompt and like, that's the third area. And the fourth area is governance and compliance.

All 4 are integrated as a single solution and you can buy some of it if you want. That's one bucket. The second is taking our exchange for agents. This is in-line policy enforcement for agents. That's a very powerful solution, highly differentiated because we're sitting in line. And that's how our customers look at it. AI Protect is already being bought significantly, exchanged in early stage working with some of the customers.

Got it. And then just how do you think about overall SASE adoption and where we are in that. I think we get questions sometimes about kind of where we are on the core business. And how do you differentiate that maybe of kind of this grab bag where SASE means a lot of things to a lot of people, so how do you differentiate kind of within those areas?

I think it's a good question because SASE whatever. So this is how you should think about it. First of all, you need to understand that Zero Trust is fundamental for security, more and more important in the world of AI. That means things A talks to B through a switchboard. That's where the world has to go. From Zscaler point of view, we have 45% of Fortune 500 companies doing that today and 40% of global 2000 companies. Our large enterprise base that we target is 20,000 companies, 4,400 of them are customers, which is less than 25%. This means we have a sizable new logo base to go after, and we are pursuing that.

Also, it means there's a lot of opportunity for upsell, okay? And we have seen a number of million dollar deals going up, number of $5 million customers going up. In fact, last quarter, we had the largest number of $1 million deals in Q2. That's excellent. We also -- we're looking at last quarter, we learned that from the initial buy -- in 4 years, our enterprise customers will have 3x of the ARR. That's pretty good. And this is over the years when our platform was much smaller. As the platform has gotten bigger, we expect that this thing should continue or get better. So both opportunities, meaning significant upsell and significant new logo.

Yes. Okay. Perfect. I want to step ahead to go to market for a second. Can you just kind of talk about the evolution of relationship with channel partners? And just how you see kind of relationships with GSIs evolving?

In relation to them?

For go-to-market.

Go-to-market, yes. So we've gone through transformation of go-to-market and we've gone through it successfully. We walked you through, we did get projections, we met our projections. We are seeing the trajectory moving in the right direction. If you look at Q2 our sales productivity was up in double digits. Our total number of large deals moving up. Our channel contributions getting much better. At a broad brush basis, when you look at lots of channel partners, you talk to them, you're going to get a mixed feedback because we don't deal with 5,000 of them. Global system integrators work well with us. We have worked very hard with them. Why do we need them?

Here's the difference. A typical VAR will sell the boxes and they may get some deployment services, may not. They largely do a lot of fulfillment. When we go in, we -- you're not replacing Zscaler with Zscaler, a firewall here, a firewall there. You are transforming, you're removing lots of stuff. To remove all these branch firewalls, routers, switches, load balance. You need somebody who can do services. GSIs do a good job there in that area. So GSIs are working well for us. Selected channel partners who do these services, they are working well for us. So it's good. Our trajectory is moving in the right direction. Now we're working on upping that work. Kevin, you want to add anything to that?

No, the only other proof point that I would just offer is we had very strong pipeline conversion in the second quarter as well. So just another indication that things are building momentum nicely.

And I know we just talked about this, Red Canary, but any -- how -- should we think of it really as getting to that -- the 25% penetration you guys just talked about. Getting that to be a higher number? Or is there anywhere down market where we should -- that you find attractive opportunities?

I mean, to Jay's point, I mean, our focus are the largest companies in the world. Our direct sales organization is targeted at enterprise customers, which we estimate to be a little more than 20,000 companies in the world. That's where we go after with direct resources. Anything below that would be served through our channel, and that would be indirect for us. So our focus really is the largest companies.

Got it. And Kevin, maybe a question for you. You guys are rule of 50 company, obviously, leading growth and profitability. How are you thinking about kind of that trajectory going forward or just opportunities for operating leverage. And I want to bring into that discussion. There's maybe been a little bit more scrutiny of stock-based compensation. Just how you guys are thinking about stock-based compensation versus cash compensation.

Yes. I mean as a tech company providing alignment to our employees and stock is important, and I think investors appreciate that alignment. We also know that it has been a little bit elevated and we are acutely focused on bringing stock-based comp down over time and achieving GAAP profitability over time. So that is certainly something that is in focus.

As I think forward in terms of areas of investment and opportunity, we expect to see continued improvement in sales efficiency as we continue to move forward. And as I've gotten through the transformation efforts that have been ongoing over with Mike joining us a couple of years ago and going through that process. So we think there's opportunities there. And not to pull on a topical thread, but AI over time, will also provide areas of leverage.

Okay. Another topical point, memory costs. You noted -- remember, you hadn't seen any impact yet, but you have a good amount of inventory, both of kind of branch appliances and data center equipment. But just how are you thinking about potential impacts to pricing and CapEx just as we work through this memory cycle?

Yes. So it's a real issue out here. If you are a real box company shipping lots of boxes, you will be feeling the pain. Since we aren't. So the impact is less because we have been maintaining good inventory during -- starting during the COVID time. But we do expect that it will have some impact and we will pass some of the cost to our customers. So we will adjust some of the pricing. And -- but I think from modeling point of view, we are still targeting to really stay at about 80% gross margin. That's a good number for us. It also allows us to really introduce products at a faster pace. Rather than trying to optimize gross margins, I would rather get in the market faster by 6 months and 9 months and take time to optimize it. But back to your pricing. It will have some impact, but not significant, and we'll manage it.

All right. Jay, maybe I want to end with, you've built an incredible company. Just how do you think about the next 3 to 5 years? And what is your vision for Zscaler over that time?

So if you look at 2 big areas, Zero Trust Everywhere is meant to do 2 things. One, make sure you don't get compromised. Number two, to make sure you don't lose data. Sometimes I jokingly tell CISOs. Your job is really 2 simple things. Nothing compromises you or breaches you and nothing leaks out. We have built the best Zero Trust Exchange, and we are in the first inning of the journey with a lot more opportunities ahead of us.

Now comes AI. AI is fundamentally changing everything at many, many levels out there. So we have some very key differentiators in AI security. The exchange is probably the biggest one for us to differentiate us. And also the amount of data, the logs, the proprietary data we are getting is probably better than anybody else. It is with SSL inspection, we can see everything. So it expands our opportunity in the SecOps area. But overall, I see our business growing in 3 pillars: Zero Trust Everywhere, data security everywhere and AI security at all angles. This will give us plenty of opportunity to get to $10 billion ARR and beyond.

All right. Perfect. Well, Jay and Kevin, thank you so much for being here today.

Thanks for having us.

Thank you. Great.

Good day, and thank you for standing by. Welcome to the Zscaler Second Quarter 2026 Earnings Call. [Operator Instructions] Please be advised that today's conference is being recorded.

I would now like to hand the conference over to your first speaker today, [ Kim Watkins ], SVP of Investor Relations and Strategic Finance. Please go ahead.

Good afternoon, and thank you for joining us today. Welcome to Zscaler's Second Quarter Fiscal 2026 Earnings Conference Call. On the call with me today are Jay Chaudhry, Chairman and CEO; and Kevin Rubin, CFO. Please note that we hosted our earnings release, shareholder letter and a supplemental financial schedule to our Investor Relations website. Unless otherwise noted, all numbers we talk about today will be on an adjusted non-GAAP basis. You will find a reconciliation of GAAP to the non-GAAP financial measures in our earnings release.

Before we get started, I'd like to remind you that today's discussion will contain forward-looking statements, including, but not limited to, the company's anticipated future revenue, annual recurring revenue, net new annual recurring revenue, gross margin, operating profit, net other income, earnings per share and free cash flow margin, our customer response to our products, our expectations regarding AI and its impact on our business and customers, and our market share and market opportunity, and our objectives and outlook. These statements and other comments are not guarantees of future performance, but rather are subject to risks and uncertainties, some of which are beyond our control. These forward-looking statements apply as of today, and you should not rely on them as representing our views in the future. We undertake no obligation to update these statements after this call. For a more complete discussion of these risks and uncertainties, please see our filings with the SEC as well as in today's earnings release.

I also want to inform you that we'll be attending the following conferences. Morgan Stanley Technology, Media and Telecom Conference on March 2. Loop Capital Markets Investor Conference on March 10. Stifel Technology Conference on March 10. Cantor Global Technology and Industrial Growth Conference on March 11, and Wells Fargo Software Symposium on April 8.

And with that, I'll turn the call over to Jay.

Thanks, Kim, and thanks to everyone for joining us today. We delivered strong Q2 results, and I couldn't be more proud of the team's execution. ARR grew 25%, reflecting continued strong demand for our platform. We are confident in our outlook for the second half of fiscal 2026. And as a result, we are increasing our guidance across the board.

I'd like to zoom out for a moment and talk about what's on everyone's mind, AI. AI is the single most transformative technology of our time and its mass adoption only just beginning. We believe Zscaler is the security platform for the AI era because we already protect users, data and applications across clouds and the Internet at scale. Just as we enable customers to securely accelerate digital transformation and cloud adoption, we believe we are uniquely positioned to secure the AI transformation driving continued demand for our platform.

Organizations are rapidly adopting AI to drive productivity and innovation. But doing so is creating new vulnerabilities, significantly expanding the attack surface, and increasing cyber threats in scale, sophistication and speed, recasting AI from a productivity engine into a dangerous security set. During my conversations with more than 100 CEOs and CIOs, including many at the World Economic Forum in Davos last month, the urgency of securing AI is one of the top concerns on their minds.

This is the opportunity for Zscaler's industry-leading Zero Trust Exchange, which enables our customers to securely scale AI for the Agentic era and beyond. Zscaler minimizes the attack surface and limits [indiscernible] with our unique Zero Trust architecture that enables direct one-to-one communication among users, applications and AI agents.

I started Zscaler with an initial focus on securing users with Zero Trust. Then we extended our platform to deliver Zero Trust for workloads, branches and devices, which has increased our TAM significantly, and extended our technology lead from other vendors who are still trying to build a SASE solution for user security. Now we are extending our global Zero Trust Exchange platform to secure AI applications, AI agent communication, and Agentic workflow at scale. As AI agents are unleashed within the enterprise, it won't be long before billions of AI agents interacting with each other will have access to mission-critical applications and sensitive data.

Just like users, in organizations AI agents are also becoming the weakest link in cybersecurity. Imagine a threat actor, highjacking even one-off and organizations AI agents, resulting in a serious breach. AI agents shift the threat landscape and operate autonomously at speeds far exceeding humans, exponentially increasing Agentic traffic, while compressing the time to prevent, detect and respond to threats. This is becoming even more acute as AI agents, or apps, exposed to the Internet can be scanned and targeted in seconds.

Securing this new reality requires in-line policy enforcement at massive scale. This is what Zscaler is built to deliver. Zscaler stands for the Zenith of scalability. Effective AI security requires a proven Global Zero Trust Exchange infrastructure. And we believe Zscaler is the only cybersecurity platform for the AI age that's able to secure at this unprecedented speed and scale, creating a durable advantage. We have 15-plus years of experience operating our own cloud across 160-plus data centers worldwide, offering real-time security services with [ 99.999% ] reliability.

This global infrastructure is critical to secure AI agent communication. Our software is present on millions of end-user devices, servers, as well as in the cloud and branch offices, and it enables us to get Agentic traffic to our Zero Trust Exchange giving us a unique advantage. With our AI platform capabilities, we processed nearly 1 trillion AI transactions in calendar 2025. We are also processing millions of MCP requests through our exchange on a monthly basis, up from literally nothing a couple of quarters ago.

As an example, a large Fortune 100 financial services customer is using our Zero Trust Exchange to enforce policy for the software development agents. In another example, a CECL of a Fortune 500 entertainment company, a Zscaler customer, shared with me that with little deployment effort, he was able to turn on Zscaler Exchange to enforce policy for AI traffic and is securing over 4 million prompts per week. Our Zero Trust Exchange is fundamentally different from competitive firewall-based security architecture that connects users or AI agents to a network, and then allows them to run free, dramatically increasing the risk of cyber breaches. We expect these advantages, including significant architectural differentiation and our large customer base to drive short-term and long-term demand for our platform.

Turning back to the quarter. Our results reflect robust demand across all 3 of our growth pillars. AI security, Zero Trust Everywhere, and Data Security Everywhere. I will start with AI security, which includes two product areas. AI Protect our recently introduced solution to secure the use of AI and Agentic operations. I will begin with AI Protect which secures the full spectrum of enterprise AI adoption and solves a range of cyber and data law challenges.

Zscaler [ threat lab ] research found that in calendar 2025, AI application use within our customer base expanded to over 3,400, a quadrupling in the last 12 months alone, with data transfers to AI apps exceeding 18,000 terabytes. Many of these apps have serious vulnerabilities. Zscaler AI Protect, gives customers a single integrated way to secure AI at scale by discovering and managing all AI assets, including shadow AI uses, enforcing safe access to approved apps and inspecting every prompt and response in real time to stop data leaks and attacks like prompt injection.

For customers building their AI models and applications, our AI [ red-teaming ] solution performs continuous security assessment. This quarter, we integrated our AI red teaming and our guardrail products to provide true closed-loop security. While our Zscaler AI Protect solution is new, we see demand rapidly accelerating, including landing new logos representing a massive future growth opportunity.

This quarter, several large enterprises adopted our solution. In an 8-figure new logo win, we landed a Fortune 500 semiconductor manufacturer. This [ comer ] expanded its use of our platform to include AI Protect and data security solutions to block access to unsanctioned applications, prevent public LLM data leakage and provide visibility into prompts. Zscaler's integrated AI Protect solutions spanning the entire AI life cycle was a key differentiator for this new logo win.

In another example, in a 7-figure upsell deal, a Global 2,000 construction company, which is securing users with Zscaler, added our AI Protect solution to prevent data leakage and enforce acceptable use controls, or access to Gen AI applications.

The second product area of our AI security is Agentic Operations, which includes our Agentic [ SecOps ] and Agentic IT Ops solutions. We are significantly advancing our Agentic SecOps capabilities by integrating Red Canary's agent framework with the deep security insights we generate from the Zscaler Zero Trust Exchange, which processes more than 500 billion transactions every day, more than 20x the number of daily Google searches. This fusion of capabilities simplifies customer operations automates threat hunting and provides more accurate actionable threat prioritization.

Some of our wins for our [ Agentic SecOp ] solution this quarter include our leading AI software and research organization, a Global 2,000 utilities energy company, and a Global 2,000 oil and gas company. In agentic IT operations, our innovations include Zscaler Digital Experience, or ZDX Copilot, which combines Agentic technology with a conversational interface to troubleshoot and resolve performance issues of applications and network and endpoint devices.

Booking for ZDX [ Advance plus ], which includes our ZDX Copilot product crossed [ $100 million ] over the last 12 months, growing more than 80% year-over-year. We are soon launching an AI agent for ZDX that will automate multiple troubleshooting tasks, resulting in faster diagnosis and resolution of performance issues. Overall, I'm very pleased to see growing demand and continued momentum for our AI security solutions.

Our next growth pillar is Zero Trust Everywhere, which includes revenue from customers who are more broadly adopting our Zero Trust architecture by purchasing all of the following. Zero Trust Users, Zero Trust [ Branch ] and Zero Trust Cloud. We pioneered the Zero Trust users market by disrupting the traditional proxy and VPN markets, and we are a clear market leader. With our Zero Trust branch, we are disrupting branch firewalls, software-defined networks on SD-WAN and MPLS networks. With Zero Trust Cloud, we are disrupting virtual firewalls in the cloud. We are seeing ARR from Zero Trust Branch and Zero Trust Cloud growing significantly as we dramatically reduced cost and complexity.

The number of Zero Trust Everywhere enterprises has grown at a rapid pace and now sits at over 550, up from over 130 a year ago. The pricing of Zero Trust Branch and Zero Trust Cloud are based upon the number of devices, the number of workloads and the amount of traffic which keeps growing. This expansion also creates a flywheel effect generating follow-on demand for our data security and AI security offerings.

Let me give an example of Zero Trust [ Brand ] win at a subsidiary of a Fortune 500 retailer, who significantly expanded its deployment of Zero Trust Branch to over 1,000 sites in a 7-figure upsell, making it one of our largest ever Zero Trust Branch deals. The use case is for this customer for frictionless M&A integration and rapidly bringing both newly acquired and greenfield sites online. This order also included our AI Protect solution to secure sensitive data from Gen AI apps.

In Q2, 45% of the total customers who bought our Zero Trust Branch solution for new logos, demonstrating that Zero Trust Branch is helping us grow on new logos. This is also a clear proof that for better cyber protection, customers want each branch to become like an Internet cafe, and [indiscernible] place SD-WAN, which is often sold by SASE vendors.

Another important part of a Zero Trust Everywhere solution is Zero Trust Cloud, which reduces cost and operational complexity by eliminating virtual firewalls in data center and cloud environments, and can be deployed in 10 minutes. We're seeing tremendous momentum for Zero Trust Cloud.

To share a customer example in one of our largest ever Zero Trust Cloud wins, a Global 2000 financial services customer signed a 7-figure deal increasing their ARR to more than [ 5 million ], up over 40%. Zero Trust Cloud is priced based on traffic, creating a natural path for ARR to grow as customer traffic grows. This deployment will eliminate a large number of virtual firewalls in their multiple cloud environments which significantly reduces the operational burden of managing firewalls in multiple clouds, and improves cyber posture by preventing lateral that movement. With significant wins, we are proving that for better cyber protection, customers want each cloud workload to become like an island, and communicate only through our Zero Trust Exchange and displace virtual firewalls. Our opportunity is to secure millions of workloads.

Our third growth pillar is Data Security Everywhere, which is 8 modules, including data discovery, data classification, posture management and data loss prevention. We are seeing significant traction driven by enterprises, consolidating the data security point products onto our integrated platform and simplifying deployments. The growing use of AI apps is making data protection essential, generating strong demand for our solution.

Let me share an example. In an 8-figure upsell win of Global 2000 financial services customer expanded its adoption of our platform by purchasing additional data security modules, strengthening protection for sensitive data across its organization. This customer selected us over well-known competitors to replace multiple legacy point products due to our unified policy for various data sources and channels. With this purchase, the ARR for this customer increased nearly 5x.

We believe Zscaler is incredibly well positioned to secure the AI era. As the number of agents expands, the unique Zero Trust architecture becomes even more crucial. Minimizing attack services and limiting lateral movement by enabling direct one-to-one communication.

In summary, our growth opportunity is straightforward. Traffic flowing through our Zero Trust Exchange for secured communication expands our revenue opportunity. In the Agentic era, the traffic from Zscalers more than 50 million users, servers, cloud workloads, branch locations and AI agents will grow exponentially, driven by billions of autonomous agents. We believe that Zero Trust communication will be the only way to apply the real-time protection customers need to adopt AI safely and securely.

We run the largest in-line globally distributed security called platform in the world, processing more than 500 billion transactions every day, more than 20x the number of daily Google searches. This proprietary anonymized data is [indiscernible] our AI engine, a powerful differentiator to stop ever-changing cleats at speed and scale. We provide the global infrastructure which enables our customers to secure communication and apply policies in real time at wire speed.

Today, we are trusted by more than 45% of Fortune 500 companies, and we expect to continue expanding our partnership over time. In addition, with just 4,400 of more than 20,000 largest enterprises in the world as Zscaler customers today, we have a significant opportunity ahead. This gives us a durable runway for long-term growth from both upsell and new logo opportunities. Protecting AI is not just a job or task for Zscaler. It is our mission. We believe Zscaler is the cybersecurity platform for the AI age.

Now I will hand it over to Kevin to walk through the financials.

Thanks, Jay. We delivered strong Q2, '26 results, exceeding our targets while investing with discipline. With 26% revenue growth and a 36% free cash flow margin, we achieved Rule of 62 performance in the first half of the year, placing us among the elite companies that consistently outperformed the Rule of 40.

Our Q2, '26 net new ARR was $156 million, up 19%, bringing total ARR to $3.4 billion, up 25% year-over-year. Net new ARR benefited from strength in large deals and volume of deals. In particular, the Americas closed twice the number of $1 million plus deals this year as compared to last year. Excluding the contribution from our acquisition of Red Canary, net new ARR was $139 million, up 7% year-over-year, and total ARR up 21%. These results compared to an exceptionally strong 24% net new ARR growth last year.

Red Canary exited Q2 with $114 million of ARR. For the first half of the year, net new ARR, excluding Red Canary, grew 10% year-over-year, accelerating from 1% last year. This quarter, our Zero Trust Internet Access, or ZIA, and Zero Trust Private Access, or ZPA, ARR remained healthy and grew in the mid-teens. We have steadily expanded our Zero Trust platform beyond users to protect branches, workloads, AI applications and now AI agents. We believe AI agents will drive a meaningful increase in machine-to-machine and agent to agent interactions over time.

In Q2, our non-seat-based metered usage solutions delivered just over 1/4 of new ACV, and the ARR tied to those offerings grew more than 100% year-over-year. Revenue of $816 million grew 26% year-over-year and 4% sequentially, exceeding the high end of our guidance. We closed Q2 with 728 customers generating over $1 million of ARR, and [ 3,886 ] customers exceeding $100,000 in ARR, both growing 18% year-over-year. We also set a record 1 million plus new ACV deals for Q2.

On a geographic basis, we saw strong growth from the Americas, which accounted for 57% of revenue, up approximately 31% year-over-year. EMEA accounted for 28% of revenue, up approximately 18%, and [ APJ ] for 15%, up approximately 23%. Remaining performance obligation, or RPO, of $6.1 billion grew approximately 31%, including approximately 47% classified as current RPO.

We are pleased with the strong execution in our account-centric sales motion, which is strengthening our position as a long-term strategic partner and driving deeper customer adoption over time. In Q2, we again delivered double-digit sales productivity growth, reflecting continued improvement in our go-to-market execution with meaningful headroom ahead. We also achieved record pipeline conversion for Q2, signaling stronger pipeline quality and improved visibility.

We continue to build strong momentum this quarter with our recently launched Z-Flex program. Z-Flex gives customers with multiyear commitments, the flexibility to activate or swap modules without starting a new procurement cycle, along with premium deployment assistance and support. This program is driving meaningful upsell, shorter sales cycles and greater forward visibility. In Q2, Z-Flex generated more than $290 million in TCV, up over 65% quarter-over-quarter. Since launching a year ago, we have delivered approximately $650 million in TCV, at an average 4-year term, underscoring customers' long-term commitment to Zscaler.

To share a couple of customer examples. In a 5-year 8-figure Z-Flex deal, a large U.S.-based finance and insurance customer nearly tripled its annual spend by expanding its module adoption across 11 existing modules, and adopting 5 new modules, including our AI security solution. In a new logo Z-Flex win, a Fortune 500 retail customer purchased 11 modules in a 5-year 8-figure deal. This customer adopted all of our Zero Trust solutions, including Zero Trust Users, Cloud and Branch, landing as a Zero Trust Everywhere customer.

Turning to M&A. I'd like to start with some color on our recent acquisitions. On February 5, we closed the acquisition of [ SquareX ], which extends Zero Trust capabilities into any browser, enabling organizations to leverage standard browsers like chrome and edge to secure access on unmanaged devices, without requiring a separate third-party enterprise browser or using outdated and costly virtual desktop infrastructure.

Next, Red Canary. On February 1, we executed the next phase of integrating the Red Canary teams with the respective Zscaler teams. Red Canary was primarily a technology and talent acquisition. As we shared when we closed this acquisition, churn or MDR businesses is higher than we experienced in our Zscaler business. Post acquisition, Red Canary's churn has been elevated. We'll be providing Red Canary ARR in Q3 and Q4.

Turning to operating performance. Non-GAAP gross margin was 80.2%, compared to 80.4% a year ago. Non-GAAP operating income of $181 million grew $41 million, or 29%, as compared to $140 million last year. Non-GAAP operating margin of 22.2% increased 50 basis points year-over-year, reflecting the sales productivity improvements I mentioned earlier, demonstrating leverage on sales and marketing.

Turning to the balance sheet. We ended the quarter with $3.5 billion in cash, cash equivalents and short-term investments and $1.7 billion of debt. In Q2, we generated $204 million in operating cash flow, up 14% year-over-year, and CapEx was $18 million or 2% of revenue. Finally, free cash flow margin was 20.7% this quarter, down from 22.1% last year, driven by the timing of cash collections.

Looking ahead I'd like to spend a minute addressing the recent increases in memory, storage and processor prices and availability. So far, we haven't seen a meaningful impact to our operations. However, it could become a factor in the future as we purchase equipment for our data centers and Zero Trust Branch appliances. We'll continue to monitor our costs and adjust customer pricing if needed.

Turning to guidance. Let me provide our outlook for Q3 and full year fiscal '26. As a reminder, these numbers are all on a non-GAAP basis. For the third quarter, we expect revenue of $834 million to $836 million, reflecting approximately 23% year-over-year growth. Gross margin of approximately 80%, operating profit of $187 million to $189 million, equating to an operating margin of 22.4% to 22.6%, net other income of approximately $25 million and earnings per share of $1 to $1.01, assuming a 21% tax rate and 167 million fully diluted shares.

For the full year fiscal 2026 ARR of $3.730 billion to $3.745 billion, or year-over-year growth of approximately 24%. This guidance implies net new ARR growth, excluding Red Canary of approximately 9.5%. For Red Canary, we expect ARR of approximately $130 million in fiscal '26, up from our prior guidance of $95 million, with net new ARR of approximately $6 million in Q3 and $10 million in Q4. This includes all the business expected in each period, including fiscal '26 renewals, upsells and new logos.

For the second half of fiscal '26, we expect approximately 40% of total net new ARR to be recognized in Q3. Revenue of $3.09 billion to $3.322 billion, reflecting year-over-year growth of 23.8% to 24.3%. We expect Red Canary revenue of approximately $125 million in fiscal '26, up from our prior guidance of $90 million. Operating profit of $742 million to $748 million, up approximately 28% to 29% year-over-year, up from our prior guidance of $732 million to $740 million. Earnings per share of $3.99 to $4.02, assuming a 21% tax rate and approximately 169 million fully diluted shares. And free cash flow margin of approximately 26.5% to 27%, reflecting CapEx in the mid-single digits as a percentage of revenue.

We are very pleased with the results we delivered in the first half of fiscal '26. We achieved 25% year-over-year ARR growth and record operating income. Excluding Red Canary, our net new ARR growth accelerated to 10% in the first half of the year, up from 1% in the same period last year. We also saw continued momentum with Z-Flex and closed a record number of $1 million plus ARR deals for Q2.

Looking ahead to the second half of the year, we believe we are well positioned to build on this momentum. We will do this by scaling our rapidly expanding AI security portfolio, expanding Zero Trust Everywhere adoption and growing our Data Security Everywhere revenue. Ultimately, we remain focused on driving durable, profitable growth with strong cash generation. I want to thank our employees, customers and partners for their continued support.

With that, operator, you may now open the call for questions. Thank you.

[Operator Instructions] Our first question will come from the line of Saket Kalia of Barclays.

Thank you, team, for the increased disclosure on Red Canary, very helpful. Jay, maybe for you. I'd love if you could talk about just the competitive backdrop a little bit. And anything you can touch on in terms of competitive win rates, and what you saw this quarter? I mean, clearly, this is a rising tide market, but there are other players as well. Maybe the question is where are you winning? And what impact, if any, are they having?

Thank you, Saket. We haven't seen much change in the competitive dynamics over the past few quarters. What we saw was a record pipeline conversion for Q2, which is wonderful. And we also had a record Q2 in terms of large deal wins and -- in Q2, and large deal wins, I mean, over $1 million.

I mean there's a fair amount of noise the market creates out there, SASE this. SASE that. SASE is a collection of all kinds of products. In many of these SASE numbers, legacy firewalls, VPNs get thrown out. But what we are seeing in the market is our customers care about, Zero Trust. And as we engage and explain Zero Trust, we almost always win. And by the way, SASE is not [indiscernible] Zero Trust, and Zero Trust is what eliminates lateral move. So very pleased with performance. Our brand has grown. Most of large enterprises like us, they know us. And I think the future is great for us.

Our next question will be coming from the line of Brad Zelnick of Deutsche Bank.

Congrats again on another great quarter, guys, and I also appreciate the additional disclosure. Kevin, it seems you're raising your full year ARR expectation by more than your overachievement in Q2, which might be from newer acquisitions? And are there any seasonal anomalies we should consider? Perhaps [indiscernible] out of Q2 or anything like that?

Thanks, Brad. I appreciate the comments and the question. First of all, just remember, our business seasonality tends to favor H2. So we are going into the second half of the year feeling confident. We do see a strong pipeline of deals going into the back half, which does give us confidence in the raise with the exception -- excluding Red Canary. So I would point to strengthen the overall business as well as just general seasonality that we see in the back half of the year.

Our next question will be coming from the line of Gregg Moskowitz of Mizuho.

Also welcome the additional disclosures. So thank you for that. Very interesting that your non-seat-based meter usage solutions are now over 25% of new ADV. That's higher than a lot of people had thought. And with the related ARR more than doubling year-over-year, this has the potential to put some upward pressure on the growth algorithm for Zscaler in the future.

But Jay, when you look deeper at these nonseed-based solutions, you gave some good color in your prepared remarks, but can you help us better understand what's really most resonating with customers today, as well as what you're most excited about going forward?

Yes. We started early on with Zscaler. For users with Zero Trust that is largely [indiscernible], but now we have Zero Trust for workloads [indiscernible] devices, and now we are extending it to AI agents as well. Even for [ users ], we did have a number of use cases that are nonseed-based. This is [indiscernible] where we are doing third-party contractors, guest WiFi or B2B data exchange with suppliers and customers.

And just on growth, on [indiscernible] has been very strong, and that's all nonuser [indiscernible] pricing. Our AI security solutions, which are starting small, but growing pretty rapidly at all nonuser-based, rather they are token-based. And yes. We are pleased to say that 1/4 of our new business came from metered usage, and we expect it to grow over time, especially with AI agents because we believe that there will be billions of AI agents. The only way to secure communication of agents is to go through Zero Trust Exchange that scales, that's highly reliable and globally distributed, and that's what we have.

And our next question will be coming from the line of Brian Essex of JPMorgan.

Another set of kudos to Kevin for the organic versus inorganic disclosure. Maybe a question for you, Jay. And we saw this quite a lot during like a decade ago when digital transformation was the buzzword, and a lot of different IT projects were classified as digital transformation products. Similarly, we're starting to hear a lot of projects where executives are throwing AI on top of their projects to get more budget. And from that perspective, are you beginning to see any attached to budgets outside of security? How are CIOs thinking about funding some of these projects and is Zscaler [indiscernible] in that?

So we are seeing CIOs trying to really move as fast as they can to implement AI security projects. And so the feeling is if I'm not doing something, I'll be left behind. That's a clear thing. I see I have talked to lots and lots of them, but they do all worry about cybersecurity. Especially when you see all these agents showing up every other week, [indiscernible] last night was for [indiscernible] and [indiscernible] and all these guys keep on coming. They are definitely creating security issues.

So our customers are asking us, what can you provide me visibility into AI assets are risk associated with that? And then start moving around, how do I control agents? How I do have a policy that can say certain agents can access or an application? Agents are somewhat like you, they're just more dangerous and they're growing at a rapid pace. So there is a high degree of interest in proper security, especially Zero Trust for agents that we provide the budget opens up, the budget either comes from the security side of it. All the CIOs are allocating some number of budget out to the AI project.

If you're spending $100 on an AI project, you spend $4, $5, $6 on security is viewed as very nominal thing. So we're not seeing budgets as an issue to do AI security projects. It does require that you need to engage at the C level. And we have very good C-level relationships, and with pretty good plan and credibility with Fortune 500 companies.

Our next question will be coming from the line of Meta Marshall of Morgan Stanley.

Maybe a question for me, kind of following up on Brian's question. Just what you're seeing in terms of sales cycles once kind of a deal is encompassing more AI? I guess just how does it change the dynamic of your kind of needing to take a more holistic view or needed to include more modules? Just what are you seeing there?

So sales cycle depends upon the scope of the project. The first thing our customers are trying to do is put their hands around what do they have in [ the AI ] environment. What public [indiscernible] being used and what private [indiscernible] are being used. So for that, we offer AI asset management, then they want to do vulnerability assessment. We've had [ teaming ] kind of stuff as they roll out the project, guardrails becomes important.

Last month, we launched a very integrated AI security [indiscernible] The sales cycle based on what modules they're doing is generally faster because they are not really trying to go after everything. They want to start somewhere, but they want an integrated solution. And by a number of customers that told me, we bought this solution from a start-up, but for 1 year till I figure out what integrated solution can I get from a trusted vendor like Zscaler who will be around for the long term? So deals sell cycles are faster. They are smaller deals to start with, and I think they'll grow over time, especially most of those deals are based on consumption or tokens. And as [indiscernible] users are [indiscernible].

And our next question will be coming from Fatima Boolani of Citi.

Kevin, this one is for you. I was hoping to take a step back to how you reconcile the comments around Red Canary seeing elevated churn, but also the close to 30% revision on your financial contribution expectation from Red Canary, both ARR and top line on revenue? So just wanted to kind of better understand, I know you sort of flagged that the Red Canary business. generally had much higher levels of churn relative to Zscaler proper. So I just kind of wanted to better understand the dichotomy between those statements. And if you can opine on that?

Yes, appreciate the question. So look, I mean, there is an element here that as we talked about when we did the acquisition. As we do secure the renewals, there is a positive impact to ARR and so you are seeing some of that come in. My commentary just around the elevated levels of renewals is just to give color around what we are seeing. As a reminder, Red Canary was a technology and talent acquisition, and it is a core feature of the Agentic stock that we are putting together and combining, and I mentioned that we moved into the next phase of our integration earlier this month, and now consolidating those teams, which we're really excited about.

So, I mean, the reconciliation is really just to give you a little sense for what we're seeing in the business and how you should think about the second half of the year.

And our next question will be coming from the line of Roger Boyd of UBS.

Jay, I wanted to touch on sales productivity. You've made a number of changes to the go-to-market strategy over the past year in order to really help guide customers towards more transformational projects. And I know you mentioned another improvement this quarter, but can you talk about kind of the future ramp you're expecting in terms of sales force productivity? Do you see further room to upside given the push towards more of these transformational deals that are bigger, but maybe more complex?

I'll give you a broader view and Kevin is going to get into more specific [indiscernible]. With the changes we've gone through, we are driving more transformation deals, better engaging with our customers. With that, you're seeing bigger deals, Z-Flex type of deals that are happening up and that's leading to improved productivity. In fact, rather we had a double-digit sales productivity growth. Very pleased with the way sales transformation as. Affluent. As we said last quarter, the transformation is done, now we keep executing [indiscernible]. Kevin?

Yes. Thanks, Jay. So just -- I want to just kind of double-click on that last point, right? So as we engage with our customers, it is -- the account-centric model is a much different level of engagement. We're seeing a lot of interest in Z-Flex and what that looks like from a strategic point of view. And so the nature of the conversation, the way in which we're engaging the larger deals that we're seeing all will lend itself to continued productivity opportunity going ahead.

So as I look forward, I would expect that we will continue to see improvement in productivity as a result. So we are seeing the benefits, and I expect that we'll continue to see an improvement over time.

And if I may add that record pipeline conversion as a good indication of that what we want to do is working record million-dollar deals in Q2, another indication of the [ results begin ].

And our next question will be coming from Ittai Kidron of Oppenheimer & Company.

Kevin, I wanted to dig into your comments on the core [ ZIA ACPA ] growth. I think you mentioned mid-teens in ARR. Can you give us a little bit more color what was that growth rate for the last 2, 3 quarters perhaps? And how do we think about expectations for your core ZIA/ZPA business for the next 2, 3 quarters?

Yes. Thanks, Ittai. I appreciate the question. We have seen a pretty consistent performance in [indiscernible] We did get some feedback that it would be helpful for you guys to get a little bit more color in that regard, which is why I added that into the script. Keep in mind that [ ZIA, ZPA ] as it relates to Zero Trust Everywhere, is the foundation and to a large degree, the base and the opportunity. If you look at the number of customers that we have today, roughly 4,400 out of more than 20,000 potential companies that we think can be customers, you look at it in terms of Fortune 500, where we still have over half of those to prospect against. There is a massive opportunity left with [indiscernible] as as we think about it.

And even within the companies that we do have on [ ZIA, ZPA ], we have an opportunity to upsell those Zero Trust Everywhere and then adjacently through the other pillars, data security and AI. So from our point of view, it just reiterates the stability in the underlying business and really gives a sense for what's driving kind of the core of the business. But again, we've got these other 3 growth pillars that have been doing exceptionally well. And hopefully, that additional color is helpful for you.

One interesting to add what [indiscernible] that customers on average are tripling their initial purchase in 4 years. That's pretty [indiscernible].

And our next question will be coming from Gray Powell of BTIG.

Great. Can you hear me, okay? I cut out there for a second.

Yes. Yes, we can. Excellent.

Okay. So I want to follow up on some of the earlier questions. I think you've hit on this somewhat. But -- so you are seeing a lot of momentum in Zero Trust Z-Flex deals. If I'm doing the math correctly, I'm calculating that Z-Flex was over 30% of RPO bookings. I'm not sure if that's how you look at it. But I guess the question is, how does the ARR ramp on a Z-Flex deal compared to customers under historical contracts? And then just any directional commentary you can give on how big a typical Z-Flex customer is at maturity versus traditional, or like what they spend? And what's sort of like -- giving you the most upside from a product perspective?

Yes. Thanks for the question. Let me maybe just orient [ it ]. I mean the way that we look at Z-Flex is it is another opportunity for us to offer a package to a customer that we think is mutually compelling because the flexibility, so they have less concern about being locked into a particular product, or product decision in the future. It gives them an opportunity to focus more on the long-term partnership versus more transactional selling in nature. And then it does give an opportunity for them to try in a much easier, less friction way, new modules that expand into those modules.

So from a -- from an offering perspective, it is a much better and more strategic way to engage. We do think over time that more and more of our customers will will adopt Z-Flex. Is not something that we mandate or push, but where we feel that it really is well positioned. The field is enabled to be able to offer Z-Flex going forward.

Your question around differences in ramps, et cetera. Fundamentally, two deals if its a Z-Flex or if its a non Z-Flex, so long as they're similar structure. There's no difference in how that shows up in ARR. [ Z-Flex ], by their nature because they're longer term, they've got more products. They may have a ramp that is built in so that the customer can deploy along their deployment plan, which could take anywhere from 6 months to a year. But I wouldn't think about Z-Flex as creating a different dynamic with respect to ARR other than it's just another level of indication that we are very strategic in that environment.

The average Z-Flex deal is typically an 8-figure TCV commitment. And for those deals that we've done thus far, it's been about a 4-year period. As we've talked about, they tend to be 3- to 5-year deals. And right now, the average is about 4. So hopefully, that's helpful color.

And our next question will be coming from the line of Jonathan Ruykhaver of Cantor Fitzgerald.

So I think, Jay, this is for you. Just curious, when you look at [ SquareX ] from my understanding you're betting browser security via an extension rather than having a dedicated secure browser. Can you just talk about that? It seems like the flexibility could be a plus, but is there any trade up between control and functionality between extension and full [indiscernible]?

And then just just curious also on your view of how critical is the browser layer to winning broader Zero Trust deals over the next couple of years?

Thank you. Very good question. So we have been offering Zero Trust isolation solution using any standard browser for managed and unmanaged devices. [indiscernible] manage no problem. [ Unmanned ] devices means they were using the standard browsers. Some customers wanted something like a device [indiscernible] on unmanaged device. And for that, One option was you buy a full loan enterprise process from a third party.

We looked at some of those acquisitions a couple of years ago. We didn't like it. Full [indiscernible] browser with its own [indiscernible] and customers don't like one more agent, or in this case, this is one more mega agent on the endpoint. So what we found was with [indiscernible] acquisition, we could add the security functionality [indiscernible] using browser extensions from unmanaged device, it's a wonderful use case, or generally for third-party type of stuff for us. So it's a clean, better solution rather than trying to have full-blown third-party browser. And it really takes cares of the gap that we have in this environment.

So we think it expands our TAM. We have lots of customers who are using browser isolation. This actually will help us expand [indiscernible] some of the third parties who will come from unmanaged devices. So very pleased with the acquisition and a fit and early market reaction [indiscernible]