AudioCodes Ltd. Aktie

Well, Marina, it's great to have you join you today for this webinar on 3 Trends Driving the Future of AI. My name is Jim Lundy. I'm the Founder and CEO at Aragon Research. And we're really going to do a deep dive into edge computing and AI and why the 2 are intertwined and why they matter. It's a very hot topic in the industry. It couldn't be a better time for this webinar, Marina. I'm really excited about this. So we have a lot to cover.

And let's just talk about the agenda, and then I'll let Marina say a few words. We're going to make this interactive, by the way. So -- and this is being recorded in case you missed it. But we're going to talk about key trends and this whole idea of what we call edge, some people call it on-premise, but we really never left edge, but it's back and it's back in a big way. Why edge infrastructure is -- there's a big rationale, including cost, particularly something called tokens.

And also that sometimes when you run AI closer to the information, you're going to do it faster. And there are some things around data privacy we'll talk about. And then some practical advice. And of course, we're going to hear a lot about what AudioCodes is up to in this area. So Marina, why don't I turn it over to you for some intro?

Jim, nice having you. It's a great pleasure having you here. My name is Marina Risher. I'm Senior Product Manager in AudioCodes, and we are welcoming everybody who has joined to the webinar. Please pay attention that you have a Q&A session, and you are welcome to write your questions during the entire webinar. We will review them and near the end we'll address them. And whatever questions we will not address, we will release a summary of the webinar with all the questions answered. So let's start.

Okay. Great. So look, so we've got some trend slides. And the idea is to kind of really get you at the end of this webinar, fully kind of aware of the kind of scenario we're looking at. I'm based in Silicon Valley. I've been analyst for a long time. We do use the technology. So this whole era of agents, and you can call them assistants, but we're kind of calling them coworkers or digital workers. It's been moving very fast, particularly in 2026.

The era of cowork arrived in February. And literally, there's a new race on. And it's -- the whole idea is that it's not just about having agents, but the idea of an assistant now, even today, things called a personal assistant that can do work for you on your behalf. And of course, one of the things that comes up when you talk about that is something called agentic identity and security. That's a hot trend. The only problem is a lot of the products for agentic identity and security are not shipping yet. So some people have a kill switch.

But understanding who the agent is or who the assistant is that is doing work on behalf of a human is very, very critical as is making sure your infrastructure is set. So we're going to talk about that. But the whole idea in 2027 is this whole idea of this full agentic digital worker that can work on your behalf and with token costs going down, can do it for a long time. So that's kind of a little bit of a teaser, but we're going to dive into all this and more. So let's go ahead and keep going.

So the return to edge computing. So there's a lot that's kind of going on behind the scenes. If you're watching the news, you're going to think that all AI runs in a hyperscaler cloud, and that's just not true. The real answer is that a lot of AI is going to run on your phone, on your computer or in your local data center. And part of the reason for that is information security, data sovereignty and cost. So there is going to be a massive demand to upgrade PCs, Windows PCs in particular, because local compute power, i.e., GPUs and inference chips, will make your machine, your local machine, your computer or your phone, an AI machine.

So if you haven't heard about that, at your enterprise, you probably need to start asking the question, particularly if you're like a managing general manager and you have employees, what is IT doing to refresh PCs? It is no longer about the cost of the PC. It's actually about the power of compute power of the PC. And there's a big race there. There's lots of different architectures, but the GPU-based approach with cores is the new way forward. Private data centers, sometimes called neoclouds, also very, very big. Massive investment.

Most data centers today, in fact, about 90% of data centers worldwide are not AI-enabled. And then the last thing, which some deals were struck this week on some nuclear power with Google is how to find power. Now there's next-generation data centers that don't use heat-seeking GPUs, but power is going to be one of the biggest issues for local data centers as well as hyperscaler data centers. So Marina, we try to define this as a quiet revolution, particularly for many regulated industries. What's your comment on that?

Well, with all this race to the computing to the AI usage in AI, which mostly goes to cloud and private data centers, there are some industries which were not meant to go to cloud from the beginning. So we are seeing in government and defense sector that the race to AI was waiting for the solution to be ready for also on-premises as AI capability that can run now locally on the private data centers with local data centers or sovereign clouds is something that now can be used also by government and defense, which were never planning to go to cloud.

Yes, I totally agree. In fact, I know a lot of work with clients and a lot of industries where they never went to cloud. So it's kind of like SaaS applications will still probably be SaaS, but there's many, many enterprise workloads where you don't want to move the data for a lot of reasons. And part of it is cost even before AI. So it's a trend. And that's why, by the way, getting a hold of a Dell GPU server is almost impossible. So that means that people are buying them, and there's a reason why they're buying them.

So the other thing that's happening, and this is really next year is robotic edge. And so it's not just fixed data centers. It's actually devices, including robots and drones and transportation vehicles that basically will need edge compute capabilities, a mini network, many cars have that already. But we think that the demand for more capacity there, more video storage, for example, for cars and trucks.

So there's actually going to be an awful lot of demand for microprocessors and for edge AI because you actually have to have the AI running on a device for it to function. In fact, 80% of computer vision applications run on a local device versus in the cloud. So that's another trend. A lot of people kind of like, it's not going to happen. But we actually do see this coming very, very fast. And I mean, for example, deliveries with the drone, mainly rural areas like Australia. But now you have -- I'm actually in Austin, Texas right now and the cybercabs, which are -- there's no driver, wildly successful, wildly successful. So it's kind of crazy, right?

And we think a lot of the robots will launch in the $20,000 to $40,000 price range for elderly health care in the next 2 years. So that's another thing that basically implies that, that's not going to be a hyperscaler data center at all. And for the business, you need to think about the ramifications of, for example, what can -- how can we automate things and what's it going to do for my company?

All right. So we're going to dive further in here talking about a lot of the stuff going on with AI assistant and agents, and it's really a preparation period. This is the time to get smart on it to understand what agents do to understand what agentic agent systems do. And just one of the things on the slide is if you just look at the weekly active users, it's crazy high, and it's going to keep going up.

But we have an example in this slide of a sales example where you have a sales assistant for an account executive and then that assistant can invoke subagents for appointments, for follow-up, for proposals and contracts. And I'm going to be honest with you, there's only a few people that are actually doing this today. Some people are actually saying, hey, we've got one sales assistant and then they're actually going to hire somebody to build these subagents. Guess what? You can actually talk to some of these studio tools now and build them yourself.

But you got to make sure they're secure. They got to make sure they don't hallucinate. So testing is -- it's like a system deployment. So this is a great use case. Marina, what do you think about the sales use case? Notes, I think meeting notes is one of the most important things in sales. And I know that's something AudioCodes does. So can you comment?

Definitely. We are in the business of meeting intelligence where what happens in the meeting is very important because it includes not only the transcript, but most important decisions, commitments, objections. And it's very important that all this information will eventually be shared among different organization members. So not only creating the summaries or action items is very important, but also make all this automatically inside your CRM system.

So as you said, Jim, the automations and the agents which are not only summarizing, but also integrating into the CRM system, the notes. So the next meeting starts from the point where you already have a summary, you already have all the points mentioned in the previous meeting and just not start with blind zero, you are sharing organizational knowledge. So all the people involved can definitely take and benefit from it.

And this shared context, which goes directly from the meeting intelligence into the CRM systems is the strongest advantage of meeting intelligence applications. So we do see the usage, and we do see the agentic being transformed from action items and summary into integration part of CRM systems.

Yes. And I would just comment the #1 issue for sales teams is notes. It's the #1 issue.

Exactly.

It's getting notes into -- yes, it's very hard to get them into the CRM.

Exactly. And the part they need to go, open the ticket, make it manually or make a summary by themselves is tedious work and they live on meetings. Sales is based on the meetings and the automation that, from the meeting, you have automatic summary, action items, decisions and that's automatically entered in the company CRM. It's a huge momentum for all those tools.

Absolutely. So yes, so there's a lot of things on the horizon. The world is changing before our eyes. It is time for you to prepare and look at all the different use cases where you can -- basically, these things are going to help to automate. And really what it's automating is knowledge. So to try to net it out, where we were before and where we're going. In the past, it was basically applications where you had to go and know what buttons to push to get information out of the application.

And now what you're seeing is this idea of something called headless. A headless experience where the human instructs the coworker assistant to do the task and the assistant goes and presses the buttons. So it's conversational, and you have to have the right connections into it. But just the example Marina mentioned about getting notes into the CRM automatically is still to this day, one of the biggest tasks we hear about because we cover revenue enablement, and it's still the #1 issue.

But there's no other issue that's bigger than where are the meeting notes. So I think you're hitting the home run there. Okay. So look, we're going to kind of transition a little bit and talk about technologies, and we have a little video we're going to try to play here to talk about something called an Aragon Research Tech Arc. So I'm going to go ahead and advance the slide and start the video.

[Presentation]

Okay, so that's a quick flyby. It does a much better job videos of explaining things than either Marina or I could do in the same amount of time, which was less than a minute. But we just highlighted some of the things in the video, agentic systems, agentic security and privacy, other things that are starting to be integrated into some of these agentic systems, including deep learning and reinforcement learning.

But again, these are -- the idea of this is that people are not waiting for the trough of disillusionment. I was at a company that started with the letter G. What we see is waves and early adopters and even middle adopters are jumping in with some of this stuff right now. They're not waiting until it's mature. So it's kind of a great way to kind of look at where you are and are we missing out? For example, if you don't have, for example, a plan for next year for agentic systems, you probably need to get one.

So anyway, so like I said, we -- these are about 45-page reports, but the reason we do videos is to make it easier to digest for -- particularly for business executives. All right. So let's keep going. We're going to talk about this whole idea of not only the edge battle, but the discussion that's going on in the industry right now about open weight versus closed models, basically open source versus closed source. And what people have been led to believe is that you must buy closed source, meaning that you're buying a subscription to maybe one of the frontier providers like OpenAI or Anthropic. But there's this other thing called open source.

And so when you start looking at right now, the big enterprises are looking at edge data centers with an open source LLM that's clean, probably made in the Western Hemisphere because, first of all, they can do the same things. It does need maybe a little bit more training. But the cost -- let's just say, if you had -- you're spending $10 to evaluate 100 documents, you might spend $1 with open source or even $0.10.

So the hyperscaler factory is generally going to be using their models. They're going to guarantee things. However, when you actually start looking at agents and if you're running their agents in their data center, which is digital labor, then a lot of the profit will actually go to the hyperscaler that's in your country. So there's actually kind of a race going on like do we want the hyperscaler in our country? Or do you want to have our own local data centers where we run the AI.

So there's a lot to unpack here, Marina. We know that and you mentioned this earlier, there's a lot of firms that have really started to wake up and realize that was the SaaS era, just an era and can we still do it on our own? And I would argue that they can. But give me your take on what you see here.

We definitely see that it's a possibility. One of the solutions we're providing today is the meeting intelligence solution, which is running fully on-premise. And as you properly said, the models are getting smaller, and you can have a smaller model, it's called SLM, small language model. And those small language models coming pretrained or we are making the training for them.

And then when it is installed on the part of the application, you are running fully the entire model and the entire solution components like STT, speech-to-text agents, LLMs, and TTS, text-to-speech components, all those components can be running locally on your premises. And we see huge importance of the ability to provide such solutions in government, in health care, in defense sector. And I will speak later a little bit about asylum interviews.

So any place where you cannot -- where data cannot leave the premises of the organization and before the edge computing era cannot -- could not benefit from AI. Today, it's possible. Today, we have solutions with models which are adaptable and can be packed up as a solution with proper scaling and fully running on premises.

Yes. It's a big race. And I think that the light bulb is starting to go on in a lot of large enterprises. But remember that with small language models, as you said, Marina, you may be able to run a lot of stuff even on local PCs, MacBooks, Mac Minis and also HP servers, Dell servers. I mean there's a lot of server providers that basically can run SLMs and actually now large language models. So you have a lot of choice.

Do not believe what the hyperscalers tell you because they're now going to say, oh, we can put stuff in your local data center. And they can, and they can do that. But you also have to look at cost. But again, a lot of times, it may be a managed service or you may run it on your own. But remember, with a command-based interface, talking to the application is much, much easier than it ever was. And you might be surprised, we've even proposed that like some field engineering teams or field SEs that they have kind of like a jam session and compete against product management about who can make the best agent. And they're like, oh, let's do that, right? I don't think the product management people like it too much.

All right. So look, this whole race, and this is just a little bit of background. It kind of talks about this whole race to build. They're calling them AI super factories. And what this chart shows is last year into this year, about basically 85% to 90% of data centers around the world were legacy, meaning they did not have AI compute capabilities. But we're looking at by 2031, a massive build-out where between 47% to 66% may have AI componentry built in in the supercompute data centers. So there's massive spending. And we think there's also going to be massive spending for local data centers.

So basically, these are new knowledge engines for economies, and you don't want to miss out. So there is a race. And the fact that hyperscale is going to give you a data center for free, it may not be the best free, right? It may not be the best. So Yes. I would just say, Marina, this is kind of like the race to who owns AI and is it the hyperscaler of the enterprise. So what's your take on this?

We think that eventually, it will be the mix. We don't think that it's black or white. We don't think -- we believe that part of the data centers will stay local and they adopt local models for AI. And companies who are using SaaS models and running in cloud will make a decision what are the workloads that they prefer to leave in cloud as a SaaS model.

And some of the data workloads will be running on-premise because it's a hybrid environment that we expect that will eventually hold for the long term, where you don't have to decide for the entire organization and for all the applications, if it's going to be fully on-premises or fully on cloud, you can make a decision per your application and per your usage. And it's not only the -- what is the LLM used and where the model is running is who is control the operation, who is the owner of the software and who is running and decided when the software is upgraded, what is the life cycle of the software. All those issues are way beyond the LLM and how to run the LLM on-prem or in cloud. So basically, we see this trend. And we think that eventually, the hybrid infrastructure will be used for the next years.

Yes, we agree. It's a great race. We also agree it will be hybrid. There will be a mix of both because there's still some things that you may want to have a partner do versus doing it all yourself. But again, the power of choice. But let's peel the onion a little bit more because one of the things that's coming up is a lot of people are railing against these AI data centers, super factories. The big shift is going to be towards custom chips or ASICs or inference computing because it takes lower power, lower water, sometimes smaller form factors.

So NVIDIA is the leader in the GPU market, but they also bought Groq, which actually makes ASICs. So they're hedging their design bet. NVIDIA is also really a big supporter of open source. So they're providing models and supporting and they just bought Hugging Face. But this is just so you know, if you're going to say, hey, I want to do a data center, you don't always have to do it with one design. The forward-looking designs are basically these inference processors. Again, these are all basically processors.

But again, there's just different design points. And you can look at the graphic on the left, that's a lot of hardware and it gets hot. It's a graphics processor. So that's one of the things you say is like how much heat does it put off. And this is the newer design. So you actually see some of the hyperscalers making their own TPUs, tensor processing units, partially due to cost and partially due to compute power. But right now, NVIDIA is the #1 choice, and they're powering everything and the GPU stuff is -- the TPU stuff is coming on gradually.

So that's just a little bit of background. So if you're going to do your own data center, you need to start looking at part 1, part 2, maybe start with GPUs, then transition down to ASICs and inferencing as supply becomes available because right now, you can't get a lot of GPUs, right, because all the hyperscalers are making them and keeping them.

Okay. So this is a real important topic that we see. And again, it's really just one of the things where you kind of have to have it. Agentic identity and security. It's tied to data privacy. But basically, agents are kind of like little kids, they can find everything, right? Agents can find everything. So if you've got stuff on your servers, on your PCs, it's going to find it. And so things like getting passwords off your systems, having passkeys and multifactor and also having agentic security and agentic identity. That's why like the personal agent thing just came online is a form of that. So it's a big deal.

And you have to kind of when you say, hey, we're going to do agents, the security stuff is not trivial. As an observation, what I would say is we work with many CISOs and they mean well. But a lot of times, the -- getting everything deployed so you have like something close to Zero Trust, they don't have that. A lot of enterprises just don't have that. They have a lot of tools. They haven't deployed them. They think one tool will do it all. So there are going to be many providers that will actually have a guarantee with their products that say, hey, we're not going to -- we won't train on your data, we won't take your data, and we have partners that will guarantee that your stuff is protected.

So that's just something to keep in mind. Marina, I don't know if you have any comments on that. I know you guys deal in a lot of regulated industries. I know that's a requirement. Do you have any comments that you have here?

Definitely. We see in the regulated industries, a clear requirement is to keep data -- some of the data should be kept on-premises because those are regulation requirements. The regulated industries have regulation requirements where data cannot leave the premises, and it's applicable for health care when privacy -- people's privacy data cannot leave the perimeter of the server of the organization. And it's applicable for government sector when you have citizen data, which also should be kept in local storages. So a lot of IDs, credit card numbers, et cetera, for government, defense and health care sector are definitely required to be kept on premises.

Absolutely. All right. So one of the things when you talk about preparing for digital labor, we think we're entering the era of role-based agents. And because just like with humans, if you're trained on a particular job, you're generally going to do that job well. There is one thing to be aware of is many times, providers are suggesting that they'll help you build an agent, but they're not giving you a packaged agent. And so we would just say that there's a higher desire to build -- buy something that's preconfigured and then you can tune it.

For example, a number of agents in sales that we talked about earlier, you're seeing in customer support, service agents, support agents. But the other thing to keep in mind is for some of the things that are just answering questions, could Agent Assist, for example, be run on a local phone on my phone. Google and Apple are opening up their architectures. They are AI devices. And so you could ask your providers to say, could you run Agent Assist for my contact center workers on the phone. So it's just something to think about when you start looking at that.

And then industry would be really the next thing is industry-based agents, for example, there's already stuff happening in legal, in health care as well, Marina. But there's regulations there. So you actually have to make sure that you have things like the security. But realize that there are many providers that do have out-of-the-box agents that may work for you in a particular role. And if you're dealing with your current incumbent SaaS provider, you need to ask them like what agents can I buy from you versus I'll build you an agent because they're all going to say I can build you an agent. But again, in some of those cases, do you want that agent to run locally, not in the cloud. So a lot of things to keep in mind as you start preparing for that.

But on the right-hand side, that just has some examples of different agents that are actually around today. A lot of agents today inside of enterprises are internally focused like internal help desk, internal HR help desk. But shopping is an area that's coming on very, very fast. The note taking that Marina is going to talk about in just a bit here is also very big.

All right. So information. Now not to get technical, but if I net out knowledge lake architecture into kind of a comment, it's about your information and where it's stored. And our argument, a lot of the big enterprises and regulators are going to keep their data and their information, their content and their documents and their images locally. And that makes it a perfect use case to apply AI locally so that you don't have to ship it up to the cloud, et cetera. It's particularly important when you start getting into image and video processing. So what it all means is that your information is very valuable. People want your information. That's why the frontier providers want you to use their tools so they can maybe borrow your information or steal it.

So what the information is going to be tuned into is both skills and answers. And so the whole idea of knowledge or information is not data, it's actually structured so that the agent can use your information to give a better answer or to automate a task faster. Now Marina, the hyperscalers has got an early lead on this with their AI factory, but we see this right now, and it's a big discussion in Silicon Valley is a war on who owns the IP. Would you comment on this?

Definitely. And you mentioned image processing, and I will also add voice. We are in the area of voice and most of the meetings are based on voice. So it's not only the data, which is outcome of the meeting like the summary, et cetera, but it's also the data is voice speaking, and this is the data which is also intellectual property. The voice itself, it's a biometric data, and it's something that sometimes companies choose to keep on premises, the transcription itself and the voice recording.

So we see that the intellectual property is both the contents of the meeting, the voice of all the participants and the outcomes of the meetings all that combined as an intellectual property. And again, in the regulated industry, some of them, it's very important that the entire data, the combination of all the outcomes of voice recording, of transcription, summary of decisions and the personal data, all of those -- all the aggregated data will be kept fully on-premises as an intellectual property of the organization.

Absolutely. So one of the things that you are looking for 2027 is think about the skills, what are the knowledge sources and what skills can you create? And remember, a lot of your HR departments already have skills databases of people, like what skills do they have? And that is something you could possibly tap into. But basically, using your information to build these skills and processes out is really one of your homework assignments for next year.

Skills are not hard to build. But remember, an agent can't do anything if you don't tell it what to do and you don't tell it how to do it. And so these are just applications. They're not -- it's not magic. It's not going to figure it out on its own. So keeping your data local, building your AI infrastructure out. But then the thing is like where is your knowledge. So that's the homework assignment for you. If you're listening to this webinar or if you're here with us live, you might have a question about that because we're going to have some Q&A here, but let's keep moving here.

All right. So some of these industries sometimes have been called high-consequence industries. I've been involved in many of these industries for a very long time. They all have their unique needs. At the end of the day, their information is highly valuable. It's highly sought after. And I know, Marina, your company, AudioCodes has dealt in a number of these regulated industries. Where do you see the most demand?

Well, we see -- from our side, we see the meeting intelligence applications. It's not meeting intelligence. It's also in health care sector. It's a meeting between doctor and the patient. In defense sector, it can be multilingual meetings where we also see not only the recording, but also translation services. And for government sector, and this is the example that I'm going to speak about is about the immigration and asylum services.

So if you could move to the next slide, please. So one of the examples and one of the agents that we have today is for multilanguage environment. It's a clear need of a lot of immigration and asylum in government agencies is to find out how to perform it live. And we're talking about the problem where we have to find the interpreter to multiple languages. The number of languages is huge. And a lot of time, the real challenge is to find the interpreter.

So one of the abilities and agents that we created with Meeting Insights application, which is also fully implemented locally is the translation, the live intelligence, live translations of the meetings where customer could have officer can speak one language and the applicant can see in live the translation in his own language. And moreover, he can also hear the translation. It's the decision of the officer during the call. And the way it goes back, applicant will speak his own language and the officer will hear and see it in the language that he understands.

So the catch here is that all this is running on-premises. And none of the data, not of the applicant and not of the entire interview process is lives in the cloud and the entire application is an add-on to the meeting intelligence, to the meeting recording and live intelligence. Moreover, it's not only the translation. A lot of time after each meeting officer have to create a summary and the ability of our application to create the summaries in clean, structured way.

So it includes both translation, summary, the predefined set of questions that were asked and answered. The decision, the reasoning and any specific customized templates, which are already built in part of the application, this entire workflow makes the knowledge base of the meeting so important and so clean and very easy to use for government offices. So we see a great demand of those applications and for agents like this in the areas of government sector.

So if you can move to the next slide, please. We can see that actually, the operational benefits are huge. The multilingual clarity is immediate. No need to find and to look for different translators. Records are structured and entire data is kept inside the organizational perimeter. So the ability to give such solution as fully on-prem solution is one of the pretty unique solutions that we have here in AudioCodes, and we support a huge number of languages, and we see this demand go and growing in different countries.

Fantastic. So look, a couple of things before we get into the summary and Q&A. This is kind of an attempt at a decision framework. And really, if you're just looking at the upper right and the lower right, basically fast decisions or very mission-critical. Those basically are an argument for edge compute. There are certainly still cloud-native things. I don't think e-mail is going to go back on-premise anytime soon. But there -- and there are hybrid candidates where you say, hey, we'll do some local and some in the cloud.

But again, you have to look at balancing this. It's a new way to look at information. It's not all SaaS. I mean the SaaS era kind of is winding down because edge is now just as important and actually sometimes much more cost effective. So you have to look at the whole formula. Look, if you just read the financial statements for people like AWS, they're not doing as many big deals because people have wised up and realized that it's not that hard to do it. And you can still have an AWS zone in your data center. So there's hybrid opportunities as well. So yes, so Marina, any comments here?

No. But I see a few interesting questions. I see a couple of interesting questions that arrived. So...

All right. Well, look, so edge computing never left. It's back. And critical knowledge and data should stay local. And again, just like cloud I/O, AI costs can surge in some environments. So you do need to understand AI costs just as you need to understand input and output costs for cloud environments. So Marina, let's dive into the Q&A.

Well, one -- first question is, what is the biggest misconception you hear from executives about moving AI to the edge?

Well, the biggest misconception is people are -- some people are saying that, oh, you can't do it locally. And this whole webinar was about the fact that you can. But that's the biggest thing that has been out there like, oh, you can't do it. And that's just not true. And hopefully, we proved that today.

Hope so. Many organizations kind of mixing up between data resiliency and sovereignty. How you would define the difference between those 2 -- combination between AI control and those 2 aspects?

Yes, good question. So look, there's one thing that's with AI is not letting people borrow or steal your information. So kind of a new term called AI sovereignty or sovereign AI. And it's tied to the data. So there's 2 levels of that is, in many cases, countries don't want their company's data to be stored in another country's data center because governments may want to spy on that. And the same thing goes for AI, which is if you don't have something like an indemnification policy from your vendor, that means that there's no guarantee they might not train on your data.

So there's -- it starts to get complicated, but the bottom line is if you don't have guarantees from your vendor about their cloud environment on your data, then don't move your data there. It's just that simple.

It's a good one because the last question that I have is actually what is the -- after deployment, if you decide to go edge and after 1 year of the edge deployment of AI, how do you decide if your investment was success or not?

Well, so it's a great point. So really, the whole idea of whether it's edge or cloud is what do you -- what are your objectives? What do you want the AI to do? Did you define it upfront? Like, for example, are we going to automate? Because most of the AI, really what it does, it's automating things. It's not -- and sometimes that means you can do it a little bit faster if it's repetitive. So if you haven't figured that out before you did it, are you going to figure it out a year later? So there are always issues with deployments.

So they may say, hey, we didn't deploy it well. We had problems, maybe we learned from that. But I would say, generally, if you define your objectives upfront, like, hey, we're going to have these 5 agents and they're going to run locally, this is what they're going to do. And you measure that each month. So you don't wait until the end of the year, you measure it each month and you make corrective actions each month. Then by the end of the year, you're going to know -- you're probably going to know it was successful or maybe, hey, maybe we could do part of that, for example, in the cloud. So yes, most of the time, what you're going to find is like, hey, there's a newer model that's cheaper, let's do that one.

That's a good outcome also a great outcome. So Jim, many thanks for the webinar for the great information you provided for the interesting presentation. And we have a few questions which are left in the Q&A section, which we will release with the recording of this webinar. Thanks for everybody who joined us and see you in the next webinar.

Thank you very much.

Thank you.

Greetings. Welcome to the AudioCodes Second Quarter 2026 Earnings Conference Call.

[Operator Instructions] Please note, this conference is being recorded.

I will now turn the conference over to your host, Roger Chuchen, Vice President of Investor Relations. You may begin.

Thank you, operator.

Hosting the call today are Shabtai Adlersberg, President and Chief Executive Officer; and Niran Baruch, Vice President of Finance and Chief Financial Officer.

Before we begin, I'd like to remind you that the information provided during this call may contain forward-looking statements relating to AudioCodes' business outlook, future economic performance, product introductions, plans and objectives related thereto.

And statements concerning assumptions made or expectations as to any future events, conditions, performance or other matters are forward-looking statements as the term is defined under U.S. federal securities laws.

Forward-looking statements are subject to various risks, uncertainties and other factors that could cause actual results to differ materially from those stated in such statements.

These risks, uncertainties and factors include, but are not limited to, the following, the effect of global economic conditions in general and conditions in AudioCodes' industry and target markets, in particular, including governmental undertakings to address such conditions, shifts in supply and demand, market acceptance of new products and the demand for existing products, the impact of competitive products and pricing on AudioCodes and its customers' products and markets.

Timely product and technology development, upgrades, the advent of artificial intelligence and the ability to manage changes in market conditions and evolving regulatory regimes as applicable, possible need for additional financing, the ability to satisfy covenants in AudioCodes financing agreements, possible impacts and disruptions from AudioCodes acquisitions, including the ability of AudioCodes to successfully integrate the products and operations of acquired companies into AudioCodes business; possible adverse impacts attributable to any pandemic or other public health crisis on our business and results of operations; the effects of the current and any future hostilities involving Israel, including in the regions in which we or our counterparties operate, which may affect our operations and may limit our ability to produce and sell our solutions.

Any disruption in our operations by the obligations of our personnel to perform military service as a result of current or future military actions involving Israel and any other factors described in AudioCodes filings made with the U.S. Securities and Exchange Commission from time to time. AudioCodes assumes no obligation to update the information.

In addition, during the call, AudioCodes will refer to non-GAAP net income and net income per share. AudioCodes has provided a full reconciliation of the non-GAAP net income and net income per share to its net income and net income per share according to GAAP in the press release that is posted on its website.

Before I turn the call over to management, I'd like to remind everyone that this call is being recorded. An archived webcast will be made available on the Investor Relations section of the company's website at the conclusion of the call.

With all that said, I'd like to turn the call over to Shabtai. Shabtai, please go ahead.

Thank you, Roger. Good morning and good afternoon, everybody. I would like to welcome all to our second quarter 2026 conference call.

With me this morning is Niran Baruch, Chief Financial Officer and Vice President of Finance of AudioCodes. Niran will start off by presenting a financial overview of the quarter. I will then review the business highlights and summary for the quarter and discuss trends and developments in our business and industry. We will then turn it into the Q&A session.

Niran?

Thank you, Shabtai, and hello, everyone.

Before I start my formal remarks, I would like to remind everyone that in conjunction with our earnings release this morning, we will post shortly on our Investor Relations website and earnings supplemental deck.

On today's call, we will be referring to both GAAP and non-GAAP financial results. The earnings press release that we issued earlier this morning contains a reconciliation of the supplemental non-GAAP financial information that I will be discussing on this call.

Revenues for the second quarter were $63 million, an increase of 3.1% over the $61.1 million reported in the second quarter of last year. Services revenues for the second quarter were $34.6 million, an increase of 6.2% over the year ago period. Services revenues in the second quarter accounted for 54.9% of total revenues.

Revenue by geographic region was as follows, North America, 50%; EMEA, 33%; Asia Pacific, 14%; and Central and Latin America, 3%. Our top 15 customers represented an aggregate of 55% of our revenues in the second quarter, of which 37% was attributed to our 10 largest distributors.

GAAP results are as follows, gross margin for the quarter was 65.7% compared to 64.1% in Q2 2025. Operating income for the second quarter was $3.2 million or 5.1% of revenues compared to operating income of $2.6 million or 4.3% of revenues in Q2 2025. Net income for the quarter was $0.5 million or $0.02 per diluted share compared to net income of $0.3 million or $0.01 per diluted share for Q2 2025.

Non-GAAP results are as follows, non-GAAP gross margin for the quarter was 65.8% compared to 64.5% in Q2 2025. Non-GAAP operating income for the second quarter was $4.6 million or 7.4% of revenues compared to $4.4 million or 7.2% of revenues in Q2 2024. Non-GAAP net income for the second quarter was $3.9 million or $0.15 per diluted share compared to $4.1 million or $0.14 per diluted share in Q2 2025.

At the end of June 2026, cash, cash equivalents, short-term bank deposits, short-term marketable securities and long-term financial investments totaled $64.2 million. Net cash provided by operating activities was $6.1 million for the second quarter of 2026. Days sales outstanding as of June 30, 2026, were 107 days.

In May 2026, we received court approval in Israel to purchase up to an aggregate amount of $25 million of additional ordinary shares. The court approval also permits us to declare a dividend of any part of this amount. The approval is valid through November 12, 2026. During the quarter, we acquired 950,000 of our ordinary shares for a total consideration of approximately $8.9 million.

Earlier this morning, we also declared a cash dividend of $0.20 per share. The aggregate amount of the dividend is approximately $4.8 million. The dividend will be paid on September 3 to all of our shareholders of record at the close of trading of August 19.

Now to provide an update on our guidance. We reiterate our guidance for non-GAAP diluted net income per share for 2026 to be in the range of $0.60 to $0.75. We are now raising our revenue guidance to a range of $251 million to $256 million compared to the previous range of $247 million to $255 million.

I will now turn the call over to Shabtai.

Thank you, Niran.

Second quarter financial results were solid. Most important, they reflect steady progress on our strategic initiative to transform and reposition AudioCodes as a valued AI-driven cloud and edge software and services company.

Our top line growth has maintained its track, driven by ongoing momentum in 2 primary growth engines, our Live Managed Services and Voice AI. Combined, these 2 units contributed to $84 million annual recurring revenue exit second quarter '26, growing 20% year-over-year and highlighting the increasing contribution of recurring high-quality revenue to our model. Annual recurring revenue has doubled in the last 3 years. That provides a strong foundation for future growth.

We delivered strong performance in the Microsoft Teams phone business. In addition to achieving 5% year-over-year growth, we saw pipeline expansion, increasing momentum in newly created opportunities and growth in the total contract value of opportunities signed.

Performance across our 2 primary segments was in line with expectations with our strong cash-generating connectivity business showing stability and Voice AI business revenue growing again over 50% year-over-year. The strong growth puts us on track to achieve our stated goal of 40% to 50% growth year-over-year for the Voice AI segment for the full year of 2026, targeting to reach $50 million by the end of 2028.

Notably, the strong momentum in Voice AI business with new opportunities, new emerging applications and delivering state-of-the-art solution ahead of competition provides us with further incentive to continue to increase investments in this area.

Now to highlights of the second quarter. Revenue growth accelerated to 3.1% year-over-year. Enterprise accounted to over 90% of revenues, led by 5% growth in the Microsoft Teams Phone business with a healthy mix of Live Managed Services bookings across Microsoft and Contact Center connectivity.

Connectivity business comprised of the business lines of gateways, SBCs and CPE has held nicely in terms of revenue. Services grew 6.2% and now represents roughly 55% of total revenues. Product revenues were about flat. Strength came from our dual growth engines, the Live family of UCC and CX, connectivity services and conversational AI.

Backlog reached nearly $90 million, up 23% from $73 million a year ago. The growing Live and Managed Services backlog converts to revenue in coming quarters, keeping top line visibility robust. Together, this dynamic supports continued Live services annual recurring revenue momentum.

Stepping back from the quarter, I'd like to spend a few minutes discussing one of the key trends shaping our industry. As generative AI continues to advance and with agentic AI emerging as key trend in recent years, we are seeing growing evidence that voice is becoming the most natural, comfortable and preferred medium for humans to interact with large language models in order to automate verbal communications.

This shift is accelerating the adoption of conversational AI across the enterprise with virtual agents and agent- assist solution emerging as some of the fastest-growing use cases. This trend further underlines our pursuit of expanding our business by adding investment in conversational AI technology and solution on top of our traditional voice connectivity business.

AudioCodes is uniquely positioned to capitalize on this opportunity. Our long-standing leadership in voice infrastructure, including gateways and session border controls, is complemented by deep expertise in orchestration and application layers that convert and connect enterprise voice environments with AI services.

This unique combination enables organizations to seamlessly integrate AI capability into existing communication environments while maintaining the reliability, security and compliance requirements of enterprise voice networks.

At the core of our Voice AI agents and agent-assist offering is VoiceAI Connect, a mature and field-proven technology and platform built through years of innovation and customer deployments. Built on our market-leading SBC technology, VoiceAI Connect enables enterprises to connect virtually any voice or telephony environment with leading cognitive voice services and both frameworks.

As innovation across the conversational AI ecosystem continues at a rapid pace, enterprises increasingly require flexibility to support heterogeneous environments consisting of both frameworks and cognitive services from various vendors.

VoiceAI Connect is uniquely positioned to address this need, as the leading pure-play voice AI orchestration platform in the market, it offers the industry's most comprehensive libraries of pre-integrated APIs and connectors and multiple deployment options, no matter whether it's deployed on-prem, private, public cloud.

This enables customers to rapidly and flexibly deploy and evolve Voice AI solution without being tied to a single AI vendor architecture. This open flexible approach aligns well with the dynamic nature of conversational AI landscape and play directly into AudioCode's strength in voice interrogability, orchestration and enterprise communication.

Now to the activity in the Microsoft space. I do remember, please, that last year, we ended with revenues of about $245 million, while revenues from the Microsoft area were about $160 million. So, this is our major business activity. And therefore, understanding how that has been evolving and developing in the second quarter does give you some idea as to the future.

So we exited 2025 with Microsoft Teams revenue above $160 million. In the second quarter of 2026, revenue grew 5% year-over-year and first half growth reached 5.6%, putting us on track for roughly $170 million by year-end, about 66% of our total company planned revenue.

Pipeline momentum was equally encouraging. New created opportunities in the space grew 14% year-over-year, pointing to a healthy demand environment ahead. Total contract value tied to Microsoft Teams activity rose 73% year-over-year to more than $20 million in the second quarter, a strong evidence of the potential in this segment.

I'm also pleased to report that a higher than $10 million plus opportunity that we won several years ago with our strongest channels in the U.S. has finally gone into production, adding several millions of those products every year over the next 3 years, meaning the level of revenue coming from connectivity, from gateways and other stuff is going substantially up with like another 3 million or 4 million every year.

Overall, the Microsoft Teams voice ecosystem remains healthy. Teams Phone still accounts for only a small share of the more than 320 million monthly active Teams users. And as voice grows more central to an AI-driven copilot workplace, we believe our long-term outlook in these markets remains favorable.

Some representative wins in the quarter include the following: one, in the area of IR education, you received follow-on purchase orders with a major state university with over 50 campuses totaling over $1 million, consisting of Live for Teams managed services, professional services and CapEx purchases for phones and video conferencing systems. This amount represents the latest wave of commitment of additional schools as part of the master agreement signed with the IT administrator.

In another area, we signed a contract with a global logistics company. In conjunction with the renewal of Teams Live Services, a long-standing global logistics customer located in Europe, has broadened its engagement by adopting our technology globally in place of traditional toll-free numbers. This expanded deployment is expected to generate meaningful cost savings and improve customer service outcomes.

Now to conversational AI activity, which was very strong in the second quarter. Glad to report that we have successfully executed our Voice AI growth strategy, delivering another quarter of more than 50% growth year-over-year. In fact, first half '26 revenue grew close to 100% over first half '25. This level of performance brings us closer to achieving our target of growing close to 50% for the overall 2026.

Our success in the conversational AI segment reflects the strength of our large enterprise customer base and our unique combination of expertise across telephony, networking, security, cloud and edge computing, collaboration technologies and AI-driven solutions.

Let me now turn to a more detailed discussion of the major business line within this segment. Let's begin with the VoiceAI Connect and Live Hub area. Getting back to it. We delivered a record-breaking quarter, driven by continued strong growth in VoiceAI Connect solution and our Live Hub self-service cloud platform.

Momentum was broad-based, supported by an accelerating pipeline, steady new logo wins and meaningful expansion across our existing customer base. Revenue growth in first half of '26 supports our target to grow this line of business over 50% year-over-year, similar to the growth achieved in previous year in 2025.

Growth in the quarter was primarily driven by significant capacity expansion from existing customers, underscoring the scalability and mission-critical nature of our platform. A key highlight was a Tier 1 health care provider in the U.S., where we have supported a steady ramp in virtual agent and agent deployments over the past 3 years.

During the quarter, we secured a purchase order that more than doubled their existing capability, driven largely by increased adoption of virtual agents. We also saw a follow-on win with a major North American retail conglomerate, which selected VoiceAI Connect to power virtual agent experiences for its primary business unit. This builds on the success we reported last quarter with one of its subsidiaries and reflects growing enterprise-wide adoption.

These wins reinforce our view that Gen AI-enabled virtual agent and agent-assist solution have reached enterprise-grade maturity and may now be entering the beginning of a broader adoption, accelerating the cycle.

To summarize this section, our pipeline continues to build with large more strategic deployments. Sales cycles for new opportunities have shortened in some cases, to approximately 6 months compared to historical range of 12 to 24 months, which points to the maturity of the overall space and existing customers are expanding capacity and increasing pace.

Next, to our AI-first contact center solution for Microsoft Teams. The product is called Voca CIC. We continue to experience strong operational momentum in the quarter, driven by new customer wins, expansion with our current existing customers and growing adoption of AI-powered solution.

Expanding presence in the financial services, following the Swiss banking Microsoft Teams contact center win highlighted last quarter, our Voca CIC contact center solution was selected by a leading Asian bank, replacing a major legacy incumbent.

This competitive win underscores the maturity of our platform and its ability to meet the stringent security, compliance and data protection requirements. It also highlights the broader opportunity in this vertical as financial services organization increasingly standardize on Microsoft Teams, creating a significant potential for Voca CIC.

We are seeing an increasing number of existing customers who originally adopted Voca CIC as part of their migration from legacy contact center to the cloud, now expand their engagement with us by adding AI capabilities. This dynamic represents a significant increase in revenue potential of each enterprise customer.

One example from the quarter, a major European airport logistics provider added our agent insights and omnichannel capabilities as part of the renewal, significantly increasing the total contract value of the engagement.

Several months ago, we launched Voca CIC AI Receptionist, a new solution targeting the small businesses. An AI-powered solution that supports multiple voice platforms, including Microsoft Teams, et cetera. Customer interest has been encouraging, particularly among organizations seeking to modernize customer interaction as part of a broader contact center strategy.

As part of this initiative, Microsoft announced Teams Phone Agent at the InfoComm 2026 trade show in June, opening Teams Phone to third-party voice agents based on Copilot. AudioCodes was the only third-party solution naming Microsoft announcement with our voice agent generally available at launch.

This recognition underscores our position as a category leader in the Microsoft Teams customer experience and our deep expertise in the Teams phone ecosystem. While AI receptionist is multi-platform by design, it goes deepest on Microsoft Teams, where our installed base and partnerships are strongest.

Short after that, Voca CIC became the first solution listed under Microsoft new Teams Voice Agent certification program. Voca CIC was already certified by Microsoft as a Team contract center on the unified integration model with the second certification, Voca CIC now holds Microsoft certification for both a Microsoft Teams contact center and an AI voice agent that sits in front of it.

As of today, we know of no other vendor that is listed with both like us. At the heart of this offering is Microsoft Teams specific version of AI receptionist delivered under Voca CIC and build end-to-end on the Microsoft technology.

Microsoft Call Automation orchestrates this interaction and Microsoft Copilot is the AI agent driving the voice agent. The entire call path runs the Microsoft stack for enterprises building their cloud and AI strategy on Microsoft that removes a lot of friction. One technology stack, one security and compliance review, one accountable provider for implementing an ongoing support under a fully managed service.

Together, these 2 milestones extend our leadership position in the Microsoft Teams customer experience category with Voca CIC. Microsoft Teams Phone continues to expand its footprint in the enterprise and every Teams phone is potentially home for both Teams native contact center and a Teams native voice agent. We expect this to translate into a stronger pipeline as enterprises move from evaluating AI voice agents to deploying them at scale.

Moving on to Meeting Insights Cloud Edition. Meeting Insight Cloud Edition maintained strong momentum during the quarter with continued growth across key operating metrics. Meeting volume was -- has doubled basically over the year ago quarter. Monthly recurring revenues grew close to 150% and the active users number reached a new record level, growing 50% year-over-year.

The sales focus of Meeting Insight until now has primarily been concentrated to 2 or 3 countries. We now plan to expand the sales activity into substantially more markets, which as you can imagine, can substantially grow revenues substantially further.

Exit second quarter 2026, Meeting Insight is a mature and proven cloud platform built through years of innovation and customer deployments. Our strategy of delivering tailored workflow solution for specific vertical markets continues to gain traction and is translating into meaningful commercial and operational momentum.

On the product innovation front, we further enhanced this platform with automatic language, multi-language operation and smart search, which enables users to query meeting insight using natural language. These capabilities improve the effectiveness of the solution to management teams and help customers extract greater value and intelligence from their meeting data.

Moving on to MIA Edge. MIA Edge is our next-generation MIA OP, MIA on-prem, which provides secured meeting analysis and intelligence extraction. MIA OP has proved market demand for Air gap off-cloud and secured meeting insights, intelligent extraction targeting mainly government and defense use cases.

We have also found much interest in this unique solution in the health care and finance sector seeking increased security. Currently, it supports cloud, on-premise, hybrid and air gap deployment models.

MIA Edge has been designed for regulated customers that cannot use AI on public clouds. It was designed to provide ultimate ownership and access control, data sovereignty and not less, the ability to control AI operation costs, which are mounting these days as a result of the ever-growing use of AI workloads by organization year-over-year.

One key effort in the second quarter and ongoing is the alignment of the solution to feed different customer environments such as Microsoft Teams, Cisco Jabber, Skype for Business, Google Meet, Zoom and Phone calls. At this stage, we have more than 25 active accounts with 10 in production and the rest in proof of concept and implementation.

And with that, I'd like to wrap up my presentation just to say that we had a good operational momentum in the second quarter of 2026, particularly with the continued strong growth for our 2 primary engines, the Live family of Managed Services and Voice AI. With the progress we are making in increasing our recurring revenues, we are on track with our target of delivering improved healthy top line growth in 2026 and beyond.

And with that, I'd like to move the call to the Q&A session. Operator?

[Operator Instructions] We have reached the end of the question-and-answer session, and I will now turn the call over to Shabtai for closing remarks.

Thank you, operator. I would like to thank everyone who attended our conference call today. With continued good business momentum in our UCaaS and CCaaS operations and continued growth in our emerging Voice AI business, we believe we are on track to continue our growth in the next coming years. We look forward to your participation in our next quarterly conference call. Thank you all. Have a nice day.

This concludes today's conference, and you may disconnect your lines at this time. Thank you for your participation.

Greetings. Welcome to the AudioCodes First Quarter 2026 Earnings Conference Call. [Operator Instructions]

Please note this conference is being recorded. I will now turn the conference over to your host, Roger Chuchen, Vice President of Investor Relations. You may begin.

Thank you, operator. Hosting the call today are Shabtai Adlersberg, President and Chief Executive Officer; and Niran Baruch, Vice President of Finance and Chief Financial Officer.

Before we begin, I'd like to remind you that the information provided during this call may contain forward-looking statements relating to AudioCodes' business outlook, future economic performance, product introductions, plans and objectives related thereto, and statements concerning assumptions made or expectations as to any future events, conditions, performance or other matters are forward-looking statements as the term is defined under U.S. securities law.

Forward-looking statements are subject to various risks, uncertainties and other factors that could cause actual results to differ materially from those stated in such statements. These risks, uncertainties and factors include, but are not limited to, the following: the effect of global economic conditions in general and conditions in AudioCodes' industry and target markets, in particular, including governmental undertakings to address such conditions, shifts in supply and demand; market acceptance of new products and the demand for existing products; the impact of competitive products and pricing on AudioCodes and its customers' products and markets; timely product and technology development upgrades, the advent of artificial intelligence and the ability to manage changes in market conditions and evolving regulatory regimes as applicable; possible need for additional financing; the ability to satisfy covenants in AudioCodes' financing agreements, possible impacts and disruptions from AudioCodes acquisitions, including the ability of AudioCodes to successfully integrate the products and operations of acquired companies into AudioCodes' business; possible adverse impacts attributable to any pandemic or other public health crisis on our business and results of operations; the effects of the current and any future hostilities involving Israel, including in the regions in which we or our counterparties operate, which may affect our operations and may limit our ability to produce and sell our solutions, any disruption in our operations by the obligations of our personnel to perform military service as a result of current or future military actions involving Israel and any other factors described in AudioCodes filings made with the U.S. Securities and Exchange Commission from time to time.

AudioCodes assumes no obligation to update the information. In addition, during the call, AudioCodes will refer to non-GAAP net income and net income per share. AudioCodes has provided a full reconciliation of the non-GAAP net income and net income per share to its net income and net income per share according to GAAP in the press release that is posted on its website.

Before I turn the call over to management, I'd like to remind everyone that this call is being recorded. An archived webcast will be made available on the Investor Relations section of the company's website at the conclusion of the call. With all that said, I'd like to turn the call over to Shabtai. Shabtai, please go ahead.

Thank you, Roger. Good morning and good afternoon, everybody. I would like to welcome all to our first quarter 2026 conference call. With me this morning is Niran Baruch, Chief Financial Officer and Vice President of Finance of AudioCodes.

Niran will start off by presenting a financial overview of the quarter. I will then review the business highlights and the summary and discuss trends and developments in our business and industry. We will then turn it into the Q&A session. Niran?

Thank you, Shabtai, and hello, everyone. Before I start my formal remarks, I would like to remind everyone that in conjunction with our earnings release this morning, we will post shortly on our Investor Relations website an earnings supplemental deck.

On today's call, we will be referring to both GAAP and non-GAAP financial results. The earnings press release that we issued earlier this morning contains a reconciliation of the supplemental non-GAAP financial information that I will be discussing on this call.

Revenues for the first quarter were $62.1 million, an increase of 2.9% over the $60.4 million reported in the first quarter of last year. Services revenues for the first quarter were $34 million, an increase of 4.3% over the year ago period.

Services revenues in the first quarter accounted for 54.7% of total revenues. Revenues by geographical region for the quarter were split as follows: North America 49%; EMEA 34%; Asia-Pacific 13%; and Central and Latin America 4%. Our top 15 customers represented an aggregate of 53% of our revenues in the first quarter, of which 34% was attributed to our eight largest distributors.

GAAP results are as follows: Gross margin for the quarter was 66.2% compared to 64.8% in Q1 2025. Operating income for the first quarter was $3.4 million or 5.4% of revenues compared to operating income of $3.6 million or 6% of revenues in Q1 2025.

Net income for the quarter was $2 million or $0.07 per diluted share compared to net income of $4 million or $0.13 per diluted share for Q1 2025.

Non-GAAP results are as follows: Non-GAAP gross margin for the quarter was 66.3% compared to 65.2% in Q1 2025. Non-GAAP operating income for the first quarter was $4.8 million or 7.7% of revenues compared to $5.4 million or 8.9% of revenues in Q1 2025.

And non-GAAP net income for the first quarter was $3.8 million or $0.14 per diluted share compared to $4.7 million or $0.15 per diluted share in Q1 2025.

At the end of March 2026, cash, cash equivalents, short-term bank deposits, short-term marketable securities and long-term financial investments totaled $68.1 million. Net cash provided by operating activities was $12.8 million for the first quarter of 2026.

Days sales outstanding as of March 31, 2026, were 104 days. On February 3, 2026, we declared a cash dividend of $0.20 per share. The dividend in aggregate amount of approximately $5.3 million was paid on March 6, 2026.

During the quarter, we acquired 1.7 million of our ordinary shares for a total consideration of approximately $13.7 million. We reiterate our guidance for revenues for 2026 to be in the range of $247 million to $255 million and non-GAAP earnings per diluted share of $0.60 to $0.75.

I will now turn the call over to Shabtai.

Thank you, Niran. I'm pleased to report solid first quarter results, reflecting continued effective execution against our strategic priorities as we continue our transformation into a voice AI-driven hybrid cloud software and services company.

Our top line growth accelerated during the quarter, driven by ongoing momentum in our two primary growth engines, our Live Managed Services and Voice AI.

Combined, these two units contributed to $80 million annual recurring revenue exit first quarter '26, growing nearly 20% year-over-year and highlighting the increasing contribution of recurring high-quality revenue to our model.

By segment, our connectivity business sustained well in the quarter, while conversational AI business grew above 50% and accounted in the first quarter for roughly 8% of revenue, underscoring the rapid uptake of our Voice AI offerings.

As discussed previously, over the past several quarters and more so in the first quarter '26, we have reallocated and increased investments in Voice AI in both R&D and sales and marketing in order to scale our channel presence and better leverage our enterprise installed base through cross-selling of value-added services.

These initiatives are clearly delivering tangible results and returns and our strong start to the year on the Voice AI puts us on track to achieve our target of 40% to 50% growth for this segment in '26 and to ultimately reach roughly $80 million of business in 2028.

First quarter growth improved to 2.9% year-over-year. Enterprise revenues accounted for over 90% of revenues in the quarter, highlighted by ongoing strength in the Microsoft business, which grew 6% year-over-year.

Overall, first quarter product revenues were about flat, while services grew 4.3% and accounted now for 55% of total revenues. Within services, the strength was driven by strong traction in our dual growth engines, namely the live family of UCaaS and CaaS, connectivity services and conversational business.

We are growing ever more optimistic about the continued strong annual recurring revenue momentum and growth prospects for the overall company, fueled by a recent next-gen live platform wins and meaningful pipeline of opportunities; and second, growing demand for productivity-enhancing GenAI value-added services.

This conviction is further reinforced by the growing backlog of Live and Managed Services that will convert to revenues in coming quarters. We exited first quarter '26 backlog with backlog at $79 million compared to $67 million from the year ago period, growth of close to 15%.

Now to our business strategy. Modern enterprise communications are highly fragmented with the organization relying on a mix of telephony, networking, security, cloud and edge computing architectures, collaboration tool like Microsoft Teams and Zoom and emerging AI-driven technologies.

As voice remains the main channel for real-time interactions, ensuring seamless, reliable, secure and compliant, integration across these diverse environments is increasingly challenging. This highlights the growing need for a unified strategy to orchestrate voice, cloud and AI application effectively and this is where AudioCodes is service.

AudioCodes utilizes a 3-layer architecture comprising infrastructure, platforms and applications to address modern voice communication and collaboration challenges. The infrastructure layer delivers secure and reliable voice communication through SBCs, gateways and devices.

The platform layer enables integration and orchestration of telephony networking, cloud communication platform and AI systems supporting environments of market leaders such as Microsoft Teams, Zoom Phone, Cisco Webex and Genesis Cloud.

The application layer provides AI-driven solutions for business outcomes, including contact center functionality, compliance analytics, recording and meeting intelligence. As such, AudioCodes is transforming from a traditional voice infrastructure provider into a leader in an AI-driven voice communication by integrating advanced voice and conversational AI technologies.

This approach enables enterprises to adopt AI solution without disrupting existing systems, reducing complexity and accelerating Voice AI adoption. This positions AudioCodes at the forefront of the evolving enterprise voice communication landscape where voice and AI are becoming increasingly interconnected.

Now to Edge Computing. Lately, cloud computing has captured most of the workload moving from premises computing. And so while cloud remains an important deployment modality, there's a growing consensus that not all workloads belong into cloud, particularly when considering data sovereignty, security, latency and cost.

This becomes even more critical as we move towards an enterprise authentic AI environment where complex multistep workflows are autonomously executed by AI systems and latency directly impacts performance and reliability.

This shift from a cloud-first or cloud-only philosophy towards a hybrid architecture optimized by use case is well-articulated in a recent report published by a leading industry analyst firm called Aragon Research.

In its report titled 2026 Edge Computing Pivot, Privacy, Control and Latency, Aragon provides in-depth analysis of edge computing as a fundamental trend shaping the future of enterprise software.

The report further highlights key verticals such as government, defense, health care and financial services as early adopters, areas that are also core targets for our meeting insights on-prem solution.

We were early in the game addressing this market need, having launched MIA OP service in Israel over 8 months ago. Today, we are in the leading -- we are a leading provider of organizational meeting intelligence for edge-based deployments.

Customer interest has accelerated meaningfully with a notable expansion in pipeline opportunities initially in Israel and increasingly across other geographies. In summary, our on-prem GenAI capabilities, combined with a broad and mature portfolio of cloud-based offering uniquely position us to capture the AI opportunity regardless of how customers choose to consume our services, cloud or edge.

Before turning to some of our business lines, let me quickly shift to our profitability metrics. As mentioned earlier, last quarter revenue totaled $62.1 million and grew 2.9% year-over-year.

Non-GAAP gross margin for the quarter of 66.3% is within our long-term target range of 65% to 68% compared to 65.2% in the first quarter '25 and 65.9% in the previous quarter. First quarter non-GAAP operating expenses of $36.4 million compared to $35 million in the first quarter -- fourth quarter of '25 and $34 million from the year ago period.

On a year-by-year basis, the higher expenses are attributable mainly to targeted investment planned to support long-term growth in the conversational AI business, our main growth engine for coming years.

In terms of workforce, we have concluded first quarter with 1,000 full-time employees, representing an increase of 2% from the 920 employees in the previous quarter and 960 employees in the year ago quarter.

Adjusted EBITDA for the quarter was $5.8 million, reflecting a 9.4% margin compared to $6.2 million or 10.2% in the year ago quarter. Non-GAAP EPS was $0.14 compared to $0.15 in the year ago quarter and in line with our plans for the year.

Net cash provided from operating activities was $12.8 million for the quarter. As you can see, we have a long list of core behind us, each generating positive cash flow.

Let's go to Microsoft highlights. First quarter Microsoft business increased 6%. This was driven by ongoing health of our live business and connectivity franchise, coupled with increasing attach rate of Voca CIC, our Teams-certified contact center solution.

Some representative wins in the quarter include the following: we signed a 48-month contract with a Tier 1 system integrator to deliver SBCs and gateways on a recurring revenue basis. The solution supports a global Teams voice deployment of a European multinational company.

Important to note that following an architectural review of the required solution, the end customer determined that its existing approach is no longer meeting its operational requirements and goals based on our assessment and recommendation, the customer transitioned to a direct routing architecture to better align with its global voice strategy.

Turning to our live platform. During first quarter, we signed a multiyear low single digit million-dollar agreement with an existing Tier 1 global care customer to transition their on-premise deployment of our services to our cloud-based service. This managed service deployment will enable this carrier to seamlessly provision connectivity service for its enterprise clients.

Finally, in first quarter '26, we recognized bookings for our initial phase of migration covering 20,000 users to the on-premise Live Pro platform for Teams voice supporting high security prison facilities in the major countries.

Upon full completion of the migration, we expect the platform to support at least 70,000 users alongside gateways, SBCs and incremental IP phone sales. Our sales team will also be looking to cross-sell our conversational AI services on top of the existing platform.

Now to Conversational AI. First quarter '26 was very successful in growing our Voice AI business. Quarterly business grew above 50% compared to the year ago quarter. We believe we are creating a strong growth engine for years to come.

Just to remind everybody that the revenue trend in that business, the Voice AI business was about $12 million in 2024, grew 40% to $16.7 million last year in 2025 and we now plan to grow by 50% and achieve $25 million at the end of this year.

Ultimately, we aim to achieve business revenue of $50 million by 2028 with strength in telephony, networking, security, cloud and edge computing, collaboration tools and AI-driven technologies. We believe we are well-positioned for growth and success in this market.

Let's now shift to a detailed discussion of each of those major business lines in the conversational AI business. Let's start first with VoiceAI Connect and Live Hub. We delivered another strong quarter, led by continued growth in our VoiceAI Connect service and our Live Hub self-service platform.

Momentum remained broad-based with steady new logo wins across the U.S., Europe and APAC, alongside meaningful expansion within our existing customer base.

Main highlights of the first quarter on the opportunity side were substantial increases of bookings, more than 80% year-over-year and steep growth in new creative opportunities of about 100% compared to the year ago quarter. So very strong uptake in bookings and newly created opportunities.

Let me mention a few notable wins. This quarter, we secured a Tier 1 win with a major North American retail conglomerate adopting VoiceAI Connect to power its virtual agent customer experience. We also see a clear path to expanding this use case into additional division.

On the Live Hub front, we continue to see encouraging traction, including traditional purchases from a multinational insurance carrier that has now tested and deployed our full suite of conversational AI capabilities, namely virtual agent, Agent Insights, IVR and code summarization.

More broadly, seeing Tier 1 enterprises adopt Live Hub underscores the strength, scalability and appeal of our all-in-one platform. Live Hub's financial performance reflects this with annual recurring revenues growing more than 20% sequentially and more than 100% year-over-year.

Overall, our VoiceAI Connect and Live Hub offerings are scaling rapidly and we are well positioned to build on this momentum as the voice agent market keeps -- continues expanding substantially in coming years.

Now to Voca CIC. We reported record invoicing in first quarter '26, growing more than 60% year-over-year. Key highlights include: first, a new contact center as a service entry in Europe, a Swiss banking institution selected Voca CIC as its exclusive platform for customer service engagement on top of Microsoft Teams, replacing its legacy contact center system.

We beat out a major Swiss contact center as a service competitor to secure this win. The selection underscores the maturity of our platform and validates its ability to meet the stringent security and data protection requirements demanded by leading banking institutions.

Extending our momentum in higher education in the U.S. was another point to mention. We further extended our leadership in North American higher education segment with the addition of another U.S. university customer who selected Voca CIC omnichannel CCaaS solution as part of a broader Microsoft Teams deployment.

This marks our 10th university customer in the region, reinforcing Voca CIC position as a trusted CCaaS provider for complex multi-stakeholder environments where Microsoft Teams is the leading ecosystem.

On the new product front, following the recent launch of Agent Insights in fourth quarter '25, our AI-driven summarization and sentiment analysis service, we successfully deployed the solution across multiple existing enterprise customers.

Early customer feedback has been highly positive, particularly around the value of custom AI-generated summaries and in surfacing actionable insight and triggering downstream CRM workflows that improve end customer outcomes. Importantly, Agent Insights represents a meaningful upsell opportunity with this service accounting for more than 50% of agency's value.

Agent Insight has been deployed with some large enterprises, including universities, airports and manufacturing facilities. Feedback so far has been extremely positive, particularly around the customer AI summary capability, which allows contact center managers to tailor and surface specific insights from customer interaction using this new generative AI-based add-on.

We identified a hot entry-level AI use case for the SMB market. We have created a stand-alone offering purely focused on the AI receptionist use case, namely providing support for automatic call routing, Q&A-based documents and web by scroll, CRM integration, appointment scheduling and outbound SMS.

Moving on to Meeting Insights Cloud Edition. Meeting Insight Cloud Edition maintained strong momentum this quarter with continued growth across key metrics. Both the number of meetings and active users again reached record levels, contributing to strong year-over-year monthly recurring revenue growth exiting March 2026.

This operational momentum was supported by ongoing product innovation. Following the extension of support with Google Meet in the fourth quarter, we expanded the platform this quarter by integrating Cisco WebEx.

With this milestone, Meeting Insight is now positioned as the go-to meeting intelligence service across all the 4 top leading UCaaS systems. We have launched new features to boost enterprise efficiency and productivity, including pre-built templates for specific roles and personas and customizable tools for business verticals.

Positive customer feedback is driving increased adoption. These value-added features, combined with our continued focus on customer workflow solutions for verticals such as higher education, municipalities, local governments, HR and finance position us well for sustained momentum in the foreseeable future.

Moving on to MIA OP. In first quarter, we experienced a significant pickup in business opportunity in both Israel and international markets with the recent geopolitical environment acting as a further catalyst to already emerging demand for edge computing.

In Israel, we signed several new customers across diverse public sector organization, each with meaningful expansion potential. We executed an agreement with one of Israel's largest health care service organization to provide transcription services for both meetings as well as customer conversation within its contact center.

We also inked an initial purchase order with the Israel national regulatory and centralized purchasing entity municipalities for municipalities. Assuming successful implementation, this customer is expected to recommend MIA OP and make it broadly available to municipal organization via its internal procurement marketplace, creating a scalable distribution channel across 200 municipalities.

Additionally, we signed a contract with a regional IDF command responsible for civilian production during emergencies. Under this engagement, MIA OP will deliver transcription and summarization services for all incoming citizen interactions, further validating our solution in mission-critical environments.

Outside of Israel, our direct sales efforts complemented by strategic channel relationships are gaining traction and driving awareness of MIA OP as a differentiated innovative solution.

As an example, we are working closely with a prominent system integrator in North America that operates a proprietary UC system serving major U.S. government agencies. Recently initiated an MIA OP proof-of-concept trial to provide meeting transcription and summarization.

Subject to successful results, we expect this relationship to serve as an entry point into broader adoption of service across large U.S. government agencies.

And with that, I'd like to wrap up my portion of the call. We had good operational momentum in the first quarter of 2026, particularly with the continued strong growth of our 2 primary engines, our live family of managed services and Voice AI.

With the progress we are making in increasing our recurring revenue, we are on track with our target of delivering improved healthy top line growth in 2026 and beyond.

And I would like to turn now the call to operator. Thank you.

[Operator Instructions] We have reached the end of the question-and-answer session, and I will now turn the call over to Shabtai for closing remarks.

Thank you, operator. I would like to thank everyone who attended our conference call today. With continued good business momentum in our UCaaS and CCaaS operation and continued growth in our emerging Voice AI business, we believe we are on track to continue growth in the next coming years.

We look forward to your participation in our next quarterly conference call. Thank you all. Have a nice day.

This concludes today's conference, and you may disconnect your lines at this time. Thank you for your participation.

Greetings, welcome to AudioCodes Fourth Quarter and Full Year 2025 Earnings Conference Call. [Operator Instructions]. Please note, this conference is being recorded. I will now turn the conference over to your host, Roger Chuchen, Vice President of Investor Relations. You may begin.

Thank you, operator. Hosting the call today are Shabtai Adlersberg, President and Chief Executive Officer; Niran Baruch, Vice President of Finance and Chief Financial Officer.

Before we begin, I'd like to remind you that the information provided during this call may contain forward-looking statements relating to AudioCodes business outlook future economic performance, product introductions, plans and objectives related thereto, and statements concerning assumptions made or expectations as to any future events, conditions, performance or other factors are forward-looking statements as the term is defined under U.S. federal securities law.

Forward-looking statements are subject to various risks, uncertainties and other factors that could cause actual results to differ materially from those stated in such statements. These risks, uncertainties and factors include, but are not limited to, the following: the effect of global economic conditions in general and conditions in AudioCodes' industry and target markets, in particular, including governmental undertakings to address such conditions, shifts in supply and demand; market acceptance of new products and the demand for existing products; the impact of competitive products and pricing on AudioCodes and its customers, products and markets; timely product and technology development, upgrades the advent of artificial intelligence and the ability to manage changes in market conditions and evolving regulatory regimes as applicable, possible need for additional financing.

The ability to satisfy [indiscernible] and AudioCodes financing agreements, possible impacts and disruptions from AudioCodes acquisitions, including the ability of AudioCodes to successfully integrate the products and operations of acquired companies into AudioCodes business; possible adverse impacts attributable to any pandemic or other public health crisis on our business and results of operations. The effects of the current and any [indiscernible] hostilities involving Israel, including in the regions in which we or our counterparties operate, which may affect our operations and may limit our ability to produce and sell our solutions; any disruption in our operations by the obligations of our personnel to perform military service as a result of current or future military actions involving Israel; and any other factors described in AudioCodes filings made with the U.S. Securities and Exchange Commission from time to time.

AudioCodes assumes no obligation to update the information. In addition, during the call, AudioCodes will refer to non-GAAP net income and income per share. AudioCodes has provided a full reconciliation of the non-GAAP net income and income per share to [indiscernible] net income and income per share according to GAAP in the press release that is posted on its website.

Before I turn the call over to management, I'd like to remind everyone that this call is being recorded. An archived webcast will be made available on the Investor Relations section of the company's website after the conclusion of the call. With all today, I'd like to turn the call over to Shabtai. Shabtai, please go ahead.

Thank you, Roger. Good morning, and good afternoon, everybody. I would like to welcome all to our fourth quarter and full year 2025 Conference Call. With me this morning is Niran Baruch, Chief Financial Officer and Vice President of Finance of AudioCodes. Niran will start off by presenting a financial overview of the quarter. I will then review the business highlights and summary for the quarter and discuss trends and developments in our business and industry. We will then turn it into the Q&A session. Niran?

Revenues for the fourth quarter were $62.6 million, an increase of 1.7% over the $61.6 million reported in the fourth quarter of last year. Full year 2025 revenues were $245.6 million, an increase of 1.4% over the $242.2 million reported in 2024. Services revenues for the fourth quarter were $34.6 million, an increase of 1% over the year ago period. Services revenues in the fourth quarter accounted for 55. 3% of total revenues. On an annual basis, service revenues were $130.7 million, an increase of 0.4% over the $130.2 million reported in 2024.

Revenues by geographical region for the quarter were split as follows: North America, 47%; EMEA, 35%; Asia Pacific, 13%; and Central and Latin America, 5%. Our top 15 customers represented an aggregate of 58% of our revenues in the fourth quarter of which 41% was attributed to our 10 largest distributors. The amount of deferred revenues as of December 31, 2025, were $84.2 million compared to $84.4 million as of December 31, 2024.

The GAAP results are as follows: gross margin for the quarter was 65.6% compared to 66.2% in Q4 2024. Operating income for the fourth quarter was $3.7 million or 6% of revenues compared to operating income of $4.1 million or 6.7% of revenues in Q4 2024.

Full year 2025 operating income was $14 million compared to operating income of $17.2 million in 2024. Net income for the quarter was $1.9 million or $0.07 per diluted share compared to net income of $6.8 million or $0.22 per diluted share for Q4 2024. Full year 2025 net income was $9 million or $0.31 per diluted share compared to $15.3 million or $0.15 per diluted share in 2024.

Non-GAAP results are as follows: non-GAAP gross margin for the quarter was 65.9% compared to 66.5% in Q4 2024. Non-GAAP operating income for the fourth quarter was $5.4 million or 8.6% of revenues compared to $7.5 million or 12.2% of revenues in Q4 2024. Full year 2025 non-GAAP operating income was $21 million compared to non-GAAP operating income of $28 million in 2024. Non-GAAP net income for the fourth quarter was $4.5 million or $0.16 per diluted share compared to $11.6 million or $0.37 per diluted share in Q4 2024.

Full year 2025 non-GAAP net income was $18.1 million or $0.61 per diluted share compared to $27.3 million or $0.87 per diluted share in 2024. At the end of December 2025, cash, cash equivalents, bank deposits, marketable securities and financial investments totaled $75.7 million. Net cash provided by operating activities was $4.1 million for the fourth quarter of 2025 and $29.4 million for the year 2025.

Days sales outstanding as of December 31, 2025, were 117 days. In October 2025, we received court approval in Israel to purchase up to an aggregate amount of $25 million of additional ordinary shares. The court approval also permits us to declare a dividend of any part of this amount. The approval is valid through April 27, 2026.

During the quarter, we acquired 667,000 of our ordinary shares for a total consideration of approximately $6.1 million. Earlier this morning, we also declared a cash dividend of $0.20 per share. The aggregate amount of the dividend is approximately $5.4 million. The dividend will be paid on March 6, 2026 to all of our shareholders of record at the close of trading of February 20, 2026.

Our guidance for the full year 2026 is as follows: we expect revenues in the range of $247 million to $255 million and non-GAAP earnings per share -- diluted earnings per share of $0.60 to $0.75. I will now turn the call over to Shabtai.

Thank you, Niran. I'm pleased to report another quarter of solid top line growth in fourth quarter '25. This performance shows our focused progress towards becoming an AI-driven hybrid cloud software and services company. 2025 marked a period of stabilization and growth for our company. After facing economic challenges in 2023 and 2024, that's affected our legacy and hardware business lines and have led to a decline in revenue in past years. We saw 2025 a recovery of our connectivity business.

Over the course of 2025, we saw promising signs of top line growth inflection. The rate of decline in legacy business has moderated, and we sell the newly invested voice AI strategic areas maintaining their robust upward trajectory. This momentum in our strategic business has been driven by our 2 primary growth engines, our live managed services and the emerging VoiceAI business.

Combined, these 2 units contributed to $79 million annual recurring revenue exit 2025, representing growth of 22% year-over-year. While holding the line in our connectivity business, we executed well on our VoiceAI initiative, growing revenues by 35% year-over-year. The transition in overall company business trajectory is a result of deliberate actions, real acting our product development investments and efforts to high market potential areas and investing in sales and marketing to build market awareness to these innovative solutions.

Looking at 2026, we plan to maintain this formula for success, improving revenue growth driving steady margin expansion and strengthening our leadership in VoiceaI driven this application for the UCaaS and [ CX ] markets. Now to highlight for our business performance in fourth quarter '25 and full year 2025. Fourth quarter total revenue grew Niran mentioned 1.7% year-over-year.

As we have continued to build on the strength of our connectivity business and successfully leveraged our enterprise customer base installed base to drive cross-sell of GenAI business voice applications that make up our conversational [indiscernible] operations. As discussed earlier, our solid fourth quarter results were marked again by strong traction in our dual growth engines, lab services delivery for UCaaS and [ CX ] and [ conversation AI ] business lines.

Specifically, and on the heels of a previous quarter, our conversational AI business increased in over 50% year-over-year for both the fourth quarter '25 and also for the second half 2025. Full year 2025 conversation AI revenues reached nearly 17 million and accounted for 7% of total revenues.

As a result, we're growing ever more optimistic about the continued [indiscernible] recurring revenue momentum for coming years. This conviction is further reinforced by the growing backlog of live and managed services that we convert to revenues in coming quarters. Exit 2025, our backlog for live services reached a level of $75 million compared to $69 million at the end of 2024.

Now let me provide more visibility into how we operate so that our overall company financial results are better understood. As stated in previous quarters, we are now in a transition period from our main focus on connectivity solution to expand and build a new AI first VoiceAI led business application operations for enterprises. I believe this will also provide more clarity into our financials still.

At this stage, our business can be generally broke down into 2 business series. The long established and running connectivity business provides for about 93% of the company's revenue. It is a mature profitable business, which runs steadily over the past 5 years and which has delivered operating margin of above 14% in 2025.

On a longer-term basis, we target this business to deliver 16% to 18% operating margin. Relying on our success in these [ missions ] along the past 10 years, we are confident in our ability to continue and drive long-term stable growth as we are the front runner in this connectivity business for both the UCaaS and the [ CX ] markets.

The second business, the VoiceAI business, focusing on Software as a Service recurring business model provided at the end of 2025, about 7% of the company revenue growing from $12 million plus in 2024 to close to $17 million exit 2025, yielding revenue growth of about 35% year-over-year. Now that several product lines reach maturity and started to produce growing annual revenue. We are confident in our ability to keep growing this business line at a right of 40% to 50% annually in coming years. and we plan to reach a revenue level of $50 million in 2028.

Need to say that we rely extensively using GenAI technology and a solution to provide business voice application for the UCaaS and [ CX ] enterprise market. It is important to note though that the VoiceAI business is in investment mode currently and generates an annual budget burn of about $9 million to $10 million a year. With the 50% annual revenue growth plan for this business line, we believe we should reach breakeven in 2 years from today.

Before turning to detailed business line discussion, let's quickly shift into the fourth quarter profitability metrics. As mentioned before, [ Foursquare ] total revenue grew 1.7%. Our non-GAAP gross margin for the quarter of 65.9% is [indiscernible] long-term target range of 65% to 68% and a slight improvement sequentially from 65.8% last quarter. Fourth quarter tariff-related cost headwinds accounted to $600,000 and aggregated to $2.7 million for the full year 2025. We expect the rate base impact to approximately get to $2.3 million in 2026.

Fourth quarter non-GAAP operating expense of $35.8 million compared to $34.7 million in the third quarter and $33.4 million from the year ago period. On a year-over-year basis, the higher expenses are attributable to targeted investment in marketing and sales tied to the VoiceAI business, allowing it to grow further and impact from the weakening U.S. dollar against the euro in the first quarter.

Full year 2025 non-GAAP operating expense increased 6.2% versus the year ago period for the same reasons. In terms of workforce, we concluded 2025 with 981 employees, representing an increase from 961 in the previous quarter and 946 at the end of 2024.

Adjusted EBITDA for the fourth quarter was $6.5 million, reflecting a 10.4% margin compared to 6.9% or 11.2% in the prior quarter. For the full year, adjusted EBITDA reached $24.8 million or 10.1% margin. Non-GAAP EPS was 16%, in line with our plans in the year ago quarter. Net cash provided by operating activities were $4.1 million for the quarter and $29.4 million for the full year 2025.

On the guidance front, we expect 2026 to be a growth year. We expect 2026 revenues of -- I'm sorry, of $247 million to $255 million in the year and non-GAAP EPS of $0.60 to $0.75. This projection assumes continued strong growth of 40% to 50% in the VoiceAI business and a stable connectivity outlook, assuming no significant changes in the macroeconomic landscape.

Our overall annual recurring revenues, which encompasses our managed services or connectivity plus conversation AI is expected to grow from 79 exit '25 on growing 20% in 2026 and reaching a range of $92 million to $98 million in '26.

Now let's move to the actual business line. Let's talk first about Microsoft. During the fourth quarter, Microsoft business saw a sequential increase of 7%. This growth was largely driven by the continued strength of the connectivity franchise and rising attach rate for [ AIFRS Avoca], which is our [ TIM-certified ] CCaaS solution. The sell contract value signing the fourth score remained consistent with previous quarters. On an annual basis, total contract value grew by 5% year-over-year, reflecting steady progress.

The Microsoft Teams voice ecosystem continues to demonstrate very healthy situation. Recently, it was disclosed that the number of [ PSTN ] users reached 26 million, up from 20 million stated in April 2024, which indicates an annual growth rate of 16% to 17%. Although Teams [indiscernible] users represent less than 10% of total Teams' monthly active worldwide user, which is estimated at 320 million [indiscernible]. There's potential of total of 80 million to 100 million pre-license users creating immediate large addressable market.

Looking at 2026, we anticipate an additional increase of 3 million to 4 million users, supporting the evolution towards an AI-powered workplaces as stated by Microsoft.

One of the [indiscernible] win was $30 million in the quarter was a 36-month contract signed with AT&T to support a large public university. [indiscernible] provides for a comprehensive range of services, including managed gateway as [indiscernible] calling plans as well as icons, facilitating the migration to Teams from Cisco. Another key contract was a 60-month deal with an international equipment manufacturer based in Europe. This engagement began with a live premium managed service for initial phase of 2,000 users, marking the start of a full migration to Teams voice from Cisco. Upon completion of migration, the focus will shift to cross-selling additional business voice applications such as [indiscernible].

In the fourth quarter, we actually -- we have been engaging in the front, extending and expanding our efforts in the U.K. market. So all the U.K. is strong. Yesterday, we announced that we now offer an end-to-end portfolio of certified voice solution for Cisco Webex calling from Cloud Connect [ PSC ] and connectivity to analog gateways and [indiscernible].

Webex calling is Cisco cloud phone system and cloud [ TV ] access provides enterprise telephony business scaling features and [ PSM ] connectivity delivered and managed to Webex cloud. For 2025, Cisco publicly stated in November that [indiscernible] ServiceNow more than 18 million users worldwide. So for us, this new evolving cooperation with Cisco represents a major new opportunity in expanding our connectivity and devices business for UCaaS in coming years.

Now to our conversational AI activity. In the last 18 months, conversational AI move from experimentation to expectation. In both UCaaS and customer experience, buyers are no longer asking should we use AI. They are asking which AI? Where does it run? Who controls the data? And how fast can we scale it. That is exactly why we have been investing in past years in developing a rich portfolio of solutions.

Across UCaaS, this is about turning conversation into business assets, meeting into decision and voice interactions into actions. Across CX, it is about moving from basic self-service bots through automation, voice engines that can resolve route, summarize, comply and improve over time.

Pivoting towards a more intelligent enterprise, our conversational AI portfolio is already built for this reality. Our solution, namely [ Voice-conect, Live hub, LocaC,Meeting Inside, Cloud Edition, meeting Insight on-prem], and more are all this [ anticonnect ] voice and conversation to enterprise systems and to support multiple models and deployment options.

But let me challenge one assumption. I see here in the market that AI value comes from the model. True, the large language model matters. However, it is a durable value that comes from orchestration, security, integration and governance. So combining our vast telephony technology base, with our conversational portfolio AI to bring your own AI approach to deploying solutions in various UCaaS and CX environments is our way to meet customers for ER, make adoption faster, reduce risk and expand what Bones can deliver. To summarize quarter, as mentioned earlier, fourth quarter '25 conversational AI revenue grew over 50% year-over-year.

Now let's start with the leading line, which is the Voice [ connect live up ] line. This discussion focuses and evolves around the conversational AI platform market and the emerging voice AI agent sector, which gained significant contraction over the past 2 years. Leading research firms estimate that the market for VoiceAI agent will reach between $8 billion to $15 billion by 2028, with expectation that it will double by 2030.

Regarding our business activities, both [ ROCA ] Connect and the [indiscernible] business delivered robust results in the fourth quarter of 2025. For the full year, this segment achieved growth exceeding 50% compared to 2024. This strong performance was driven by consistent acquisition of new clients across the U.S., Europe and APAC, as well as considerable expansion within our existing customer base.

Live hub service our voice CPaaS self-service cloud platform, empowering voice board developers to build solutions such as conversation [ IVR], VoiceAI agents, agent assist and real-time translation services. In late third quarter 2025, we announced enhancement to the [ Liva Posta ] offering notably the integration of newly developed voice agents. By year-end 2025, Live app experienced a substantial increase in both number of developers and platform usage in minutes, while monthly recurring revenue approach 150% increase compared to the fourth quarter in 2024.

Notably, many existing VoiceAI Connect [ Live Hub ] customers have accelerated their consumption rates beyond the initial projections, reinforcing our belief that the adoption of GenAI enabled an agent assist application is entering a phase of rapid growth.

A significant achievement in fourth quarter '25 was securing an initial order with a Tier 1 international carrier adopting our VoiceAI connect service to support their consumerization solution. The deployment initially targets enterprise fixed line customers, with plans to expand to the entire mobile consumer and enterprise user base in late 2026. We view this contract as an important entry point with substantial potential for further expansion as the service is scaled across current clients and new use cases are developed.

Now to Voca CIC. Voca CIC, which targeted sub-1,000 agent range recorded another quarter of strong revenue growth for both fourth quarter and full year. Revenue for the year grew over 55% compared to previous year. 2025 was very proactive in terms of progress in the Voca business line. During the year, we have developed cooperation with regional channel partners as well as with global system integrators.

Activity has been fairly positive. By now, Voca CIC has more than 200 enterprise customers worldwide. We saw extremely successful -- we are extremely successful in the education space, especially in North America, U.K. and other regions, where Microsoft Teams is dominant in the vertical. We have now more than 15 universities accounts acquired in 2025. We have introduced new out-of-the-box practical AI experiences, such as agency insights and AI receptionist, some of which extends beyond the Microsoft Teams install cell based. We have productizing on-prem survival version of Voca CIC to act as a backup in case of cloud outage.

Q4 quarter highlights include extending our momentum in higher education markets, not only in the U.S. but also outside the U.S. We can talk about a large university in South Africa, selected a CIC contact center as part of their overall Microsoft teams, you see CX deployment, success in other major winning successful large-scale enterprise deployment with a top 5 global BPO provider.

During the quarter, we issued a press release highlighting the deployment of Voca CIC with [ Atento ] on a deal, one in the prior quarter. The new conversational AI voice solution supports more than 500 concurrent AI voice agent for a large stare organization and was delivered in just few weeks compared to a typical 3- to 6-month deployment time line for a project of this scale.

New product was introduced [ Agent Insights], as discussed earlier, last quarter. We recently launched [ Agent Infant], which brings GenAI into the Voca CIC platform. Argentine side provides contact center with customizable AI summaries, sentiment analysis and one-click CRM update built natively in agent workflow.

Looking ahead, we expect 2026 to be another year of strong revenue growth, driven by continued traction in both direct sales and channel partnerships. Moving on to Meeting Insights, Cloud Edition. Meeting [indiscernible] had cloud addition maintained impressive momentum throughout this quarter, seeing consistent increases in new customer acquisitions. Record numbers were again achieved in metrics such as general meetings and unique active users, leading to substantial year-over-year, mostly recurring revenue growth as of December 2025.

This strong performance was driven by continued product innovation, posting demand both across wider markets and within customer workflow solutions designed for specific verticals such as [indiscernible], local governments HR, finance and more. Meeting Insights now works independently of any particular use systems, expanding its flexibility. In fourth quarter 2025, support was added for Google Meet and we do expect integration with Cisco WebEx in the current quarter, that adding to the existing compatibility we have with Microsoft Teams and Zoom. This updates enable GenAI meeting summaries for interactions on a major UC platform, so less in-person meetings.

Beyond allowing customers to customize prompt for their precise requirements. The platform now offers prebuilt templates created for specific enterprise roles in [indiscernible], including those legal and HR. This feature is expected to further streamline how efficiently customers can extract use in [indiscernible] meetings. Additionally, the platform's model app enables on-demand recording, action [indiscernible] management, meeting preparations and chat-based search of meeting records. These features make the meeting [indiscernible] mobile app essential for daily office operations.

Now to another derivative of the Meeting Insights solution, which we call [indiscernible], [ EMEA on-prem]. Let's talk first about the cloud repatriation trend emerging. The proportion of businesses planning to retain users on-premis jumped from 5% to 15% over 2 years, driven primarily by data sovereignty concerns in European markets and regulated sectors such as legal, finance and defense.

Even [ cloud ] committed the enterprise now scrutinize where data is hosted and processed. This trends validates hybrid deployment capabilities and position data residency controls is competitive differentiators, countering pure cloud narratives that dominate previous market cycles. In fourth quarter '25, we continue to make good progress with newly introduced [ MOP ] solution with a growing number of wins in the government and defense market in Israel, positioning the business line to account for a meaningful growth in our conversational segment in 2026.

Following last quarter, Israeli [ Nimbus ] contract award, which streamlines procurement for meeting intelligent services for all Israel government ministries and agencies, we have already signed 1 first deal and currently, I have several more additional proof-of-concept engagements across various ministries.

We also received [ Nibor ] approval certifying that our solution meets the highest standard of security and compliance standards under the [ Nimbus ] Israel and [indiscernible] framework. We expect this designation to expand both the number of agencies, we can serve and the range of services we can provide.

The [ MLP ] solution suppose currently English and the U.S. English and Hebrew in languages. We expect to substantially grow that number of supported languages to 10 basically already in this first quarter. So we expect deployment of [ MLP ] in more countries already in the second quarter and beyond.

So to wrap up my presentation, we exit 2025 with good operational momentum. The connectivity business has stabilized in second half of the year. VoiceAI business grew 35% on a yearly basis and about 50% in the second half of the year with the continued pace of investment in our live managed services activity, and in the VoiceAI area, we expect continued momentum in 2026 and beyond. And I'd like to move over the call to the Q&A session. Thank you.

[Operator Instructions] Your first question for today is from Joshua Reilly with Needham & Company.

Maybe just starting off on the updated financial target for conversational AI growth through 2028. Is the 40% to 50% annual growth, is that intended to be a CAGR growth rate through 2028? And then along with that, should we think about the primary driver being customer growth or higher spend per customer driving that conversational AI growth? So do you expect to get a lot more new customers or sell more of the new conversational AI products to existing customers?

Right. Thank you, Josh. Yes, actually, we're looking for both. As I've mentioned before, several for conversational AI business voice application reach maturity in 2025, which really says that we just started out with a few hundreds of customers. We expect that this number will grow substantially as we're adding more capabilities and more features and also investing our sales operations. I would say that in 2025, our sales ability was a bit restrained simply because we didn't want to move too quickly into the sales phase without having a more mature, more complete product.

Now we feel fairly confident with and we get the feedback from customers. So yes, the number of potential customers should grow, I would say -- I'll use the word, I'm not using usually, it will grow dramatically, I expect in certain areas. Also, the utilization of new capabilities and new features, we do expect that per customer expand on our solution will grow simply because we intend to bring more capability.

So yes, growth should come from both. And I'll tell you that I'm talking now about 50% growth. But as we talk, there are new applications popping up on a weekly basis, talking to customers. And again, our ability to combine our vast telephony capabilities with the very large investment we made in conversation AI, just to give you a data point. We are known to be a company that's investing -- reaching R&D out of about 1,000 employees. We have 350 employees doing R&D work. We are moving fastly into moving big portion of that R&D force into VoiceAI. So when we started out back in [ 2011], we had only about 40 to 50 employees on this [indiscernible]. Now we have 150 out of those 350 employees. So all in all, big investment, we see success that gives us [indiscernible] to continue to invest and believe in growth such as 50% and could be more.

Got you. And then you mentioned there has been a shift in market expectations around AI. Can you just help us understand how has that maybe positively impacted your pipeline visibility in size now that we're moving past the kind of a testing phase for customers with some of these VoiceAI products and now moving into broader adoption do you feel that your pipeline visibility and size is improving and increasing?

Yes. As I've mentioned, we are increasing our sales [ fourth], we're spreading our operation into more countries. Some of these applications are fairly easy to use, SaaS applications that a company can test, do a proof of concept for 30 to 60 days and then moving into production.

And with some of the more compelling capabilities we're bringing to the game, we do have better visibility compared to take networking deals or connectivity deals being larger, but still usually takes substantially more time could turn to be anywhere between 3 months to 9 months.

Got you. And then last question for me is, how should we think about any impact from tariffs to the 2026 financials and gross margin and any other items to be considering regarding tariffs in 2026?

Right. So gross margin, we believe, will step up simply because our product -- mix of products will turn substantially more into software and services. So we do expect to keep that range of 65% to 68% operating margin -- I'm sorry, gross margins and gross margin. And then -- I'm sorry, what was the second one? Yes. The tariff was about $2.7 million in 2025. We currently estimate it to be a bit lower, probably around $2.3 million in '26.

[Operator Instructions]. We have reached the end of the question-and-answer session, and I will now turn it over to Shabtai for closing remarks.

Thank you, operator. I'd like to thank everyone who attended our conference call today. With continued good business momentum in our live [ managed ] services operations and continued growth in our VoiceAI business, we believe we are on track to grow revenue and profitability in the next coming years. We look forward to your participation in our next quarterly conference call. Thank you all. Have a nice day.

Thank you. This concludes today's conference, and you may disconnect your lines at this time. Thank you for your participation.

Greetings, and welcome to the AudioCodes Third Quarter 2025 Earnings Conference Call. [Operator Instructions] And please note, this conference is being recorded.

I will now turn the conference over to your host, Mr. Roger Chuchen, Vice President of Investor Relations. Sir, the floor is yours.

Thank you, operator. Hosting the call today are Shabtai Adlersberg, President and Chief Executive Officer; and Niran Baruch, Vice President of Finance and Chief Financial Officer. Before we begin, I'd like to remind you that the information provided during this call may contain forward-looking statements relating to AudioCodes' business outlook, future economic performance, product introductions, plans and objectives related thereto, and statements concerning assumptions made or expectations as to any future events, conditions, performance or other matters are forward-looking statements as the term is defined under U.S. federal securities law.

Forward-looking statements are subject to various risks, uncertainties and other factors that could cause actual results to differ materially from those stated in such statements. These risks, uncertainties and factors include, but are not limited to, the following: the effect of global economic conditions in general and conditions in AudioCodes' industry and target markets, in particular, including governmental undertakings to address such conditions, shifts in supply and demand, market acceptance of new products and the demand for existing products, the impact of competitive products and pricing on AudioCodes and its customers' products and markets; timely product and technology development upgrades the event of artificial intelligence and the ability to manage changes in market conditions and evolving regulatory regimes as applicable, possible need for additional financing; the ability to satisfy covenants in AudioCodes financing agreements, possible impacts and disruptions from AudioCodes acquisitions, including the ability of AudioCodes to successfully integrate the products and operations of acquired companies into AudioCodes business; possible adverse impacts attributable to any pandemic or other public health crisis on our business and results of operations; the effects of the current and any future hostilities involving Israel, including in the regions in which we or our counterparties operate, which may affect our operations and may limit our ability to produce and sell our solutions, any disruption in our operations by the obligations of our personnel to perform military service as a result of current or future military actions involving Israel and any other factors described in AudioCodes' filings made with the U.S. Securities and Exchange Commission from time to time.

AudioCodes assumes no obligation to update the information. In addition, during the call, AudioCodes will refer to non-GAAP net income and net income per share. AudioCodes has provided a full reconciliation of the non-GAAP net income and net income per share to its net income and net income per share according to GAAP in the press release that is posted on its website.

Before I turn the call over to management, I'd like to remind everyone that this call is being recorded, and an archived webcast will be made available on the Investor Relations section of the company's website at the conclusion of the call. With all that said, I'd like to turn the call over to Shabtai. Shabtai, please go ahead.

Thank you, Roger. Good morning and good afternoon, everybody. I would like to welcome all to our third quarter 2025 conference call. With me this morning is Niran Baruch, Chief Financial Officer and Vice President of Finance at AudioCodes. Niran will start off by presenting a financial overview of the quarter. I will then review the business highlights and summary for the quarter and discuss trends and developments in our business and industry. We will then turn it into the Q&A session.

Niran?

Thank you, Shabtai, and hello, everyone. Before I start my formal remarks, I would like to remind everyone that in conjunction with our earnings release this morning, we will post shortly on our Investor Relations website an earnings supplemental deck. On today's call, we will be referring to both GAAP and non-GAAP financial results. The earnings press release that we issued earlier this morning contains a reconciliation of the supplemental non-GAAP financial information that I will be discussing on this call.

Revenues for the third quarter were $61.5 million, an increase of 2.2% over the $60.2 million reported in the third quarter of last year. Services revenues for the quarter were $30.9 million, a decrease of 4.8% over a year ago period. Services revenues in the third quarter accounted for 50.3% of total revenues. The amount of deferred revenues as of September 30, 2025, was $81.6 million compared to $78.6 million as of September 30, 2024.

Revenues by geographical region for the quarter were split as follow: North America, 48%; EMEA, 33%; Asia Pacific, 15%; and Central and Latin America, 4%. Our top 15 customers represented an aggregate of 53% of our revenues in the third quarter, of which 38% was attributed to our 10 largest distributors.

In the third quarter of 2025, we experienced increased expenses due to the implementation of the new tariff of U.S. imports accounting to approximately $0.5 million additional cost, which impacted on both GAAP and non-GAAP. GAAP results are as follows. Gross margin for the quarter was 65.5% compared to 65.2% in Q3 2024. Operating income for the third quarter was $4.1 million or 6.6% of revenues compared to operating income of $4.9 million or 8.1% of revenues in Q3 2024. EBITDA for the quarter was $5.2 million compared to EBITDA of $5.9 million for Q3 2024.

Net income for the quarter was $2.7 million or $0.10 per diluted share, compared to net income of $2.7 million or $0.09 per diluted share for Q3 2024. Non-GAAP results are as follow. Non-GAAP gross margin for the quarter was 65.8% compared to 65.6% in Q3 2024. Non-GAAP operating income for the third quarter was $5.8 million or 9.5% of revenues compared to $7 million or 11.7% of revenues in Q3 2024.

Non-GAAP EBITDA for the quarter was $6.9 million compared to non-GAAP EBITDA of $7.9 million for Q3 2024. Non-GAAP net income for the third quarter was $4.9 million or $0.17 per diluted share compared to $4.9 million or $0.16 per diluted share in Q3 2024. At the end of September 2025, cash, cash equivalents, bank deposits, marketable securities and financial investment totaled $79.7 million.

Net cash provided by operating activities was $4.1 million for the third quarter of 2025. Days sales outstanding as of September 30, 2025, were 122 days. In July 2025, we received court approval in Israel to purchase up to an aggregate amount of $25 million of additional ordinary shares. The court approval also permit us to declare a dividend of any part of this amount. The approval is valid through December 30, 2025.

On July 29, 2025, we declared a cash dividend of $0.20 per share. The aggregate amount of the dividend was approximately $5.6 million. The dividend was paid on August 28, 2025, to our shareholders of record at the close of trading on August 14, 2025. During the quarter, we acquired 1,267,000 of our ordinary shares for a total consideration of approximately $12.7 million.

Regarding the direct cost impact from the tariff announced since the beginning of 2025, we expect roughly $3 million of cost burden for the full year 2025. Given the recent stabilization in the tariff developments, we are resuming our practice of providing full year outlook. For 2025, we expect revenues of $244 million to $246 million and non-GAAP earning per share of $0.60 to $0.64.

I will now turn the call over to Shabtai.

Thank you, Niran. I'm pleased to report solid third consecutive quarter of top line growth in the third quarter and execution for our strategic objectives amidst our long-term transformation to an AI-driven hybrid cloud software and services company. In the quarter, we continued to build on the strength of our UCaaS and CCaaS connectivity business, accounting now for over 90% of our revenue and successfully leveraged our enterprise customer base to drive cross-sell of our fast-growing GenAI business applications that make up our Conversational AI division.

In fact, in many ways, we can say that as of now, AudioCode has put Voice AI front and center going forward in our operations in terms of sustained growth. Our solid third quarter results were marked by strong traction in our dual growth engines, namely the Live family of Unified Communication and Collaboration and customer experience connectivity services and conversational AI business line. In fact, our conversational AI business increased 50% in the quarter, putting us on track to reach the 40% to 50% growth for the full year 2025.

Together, these 2 units drove our annual recurring revenue exit third quarter to $75 million or up 25% year-over-year, which positioned us to reach our full year target of $78 million to $82 million. We are growing ever more optimistic about the continued strong ARR momentum and growth prospect for the overall company, fueled by a strong pipeline of opportunities catalyzed by recent launch of the next-gen live platform and the growing demand for productivity-enhancing GenAI value-add services.

This is further reinforced by the growing backlog of live and managed services that will convert to revenue in the coming quarters. We ended third quarter backlog at $76 million, growing 13.4% over the year ago backlog of $67 million.

Let me share some key developments in our strategic business lines that underscore our growing confidence in our growth prospect. We have seen growing demand from partners for our live platform, an all-in-one cloud software stack that empowers them to seamlessly integrate connectivity with GenAI-powered business voice applications. To that end, in the third quarter, we signed a live platform agreement with a global Tier 1 system integrator. This strategic landmark deal calls for alignment and coordination of all sales aspects from initial opportunity pursuit to post-sales delivery.

This comprehensive approach ensures customer satisfaction and success. The initial scope of the agreement provides managed SVC and Gateway as a service in support of major UC and CX platforms for greenfield deployments and for existing customers looking to transition their legacy infrastructure to the cloud. Where applicable, the partner will also cross-sell our award-winning Teams certified Voca Contact center, delivering a unified UCCX experience.

Based on the currently committed services, we anticipated low single-digit millions in recurring revenue during the first year of operation in this agreement. This strategic agreement represents a clear win-win for both parties. For the Tier 1 system integrator, our all-in-one UCCX conversational AI stack simplifies operations, reduces cost to serve and enhances end customer experience. For us, it significantly expands our market reach and scales our go-to-market execution in the enterprise space.

Together with our long-standing successful partnership with AT&T in North America, this announcement reinforces our market credibility and position us as a partner of choice for all AI-infused UCCX services. We are seeing strong interest from other Tier 1 prospects. Other Tier 1 system integrator prospects recognize the transformative potential and cost efficiencies of our integrated platform. We look forward to sharing additional updates on new partnership with global system integrators in coming quarters.

Now to conversational AI. In addition to pull-through of conversational AI from live platform partners, we are seeing broad-based interest in our Gen AI-powered voice application from end customers. Specifically, I would like to highlight the progress we are making in our newer service, Meeting Insights on-prem, which we call Mia OP. This is our unique Gen AI-powered meeting intelligence platform, providing transcription, summarization, automation and connectivity to other leading enterprise IT application that is completely detached from the Internet and that is tailored for regulated and security-sensitive industries.

Launched earlier this year, we have gained significant traction in the Israeli market, mainly in the government space, all through word of mouth. Recently, our leading position in the Israeli market were further cemented when we were officially awarded a contract under Project Nimbus, the Israeli government's multiyear cloud migration initiative.

As the exclusive provider of meeting intelligence services in the non-SaaS category for calendar year 2026, this award streamlines procurement for all Israeli agencies, both civilian and military, allowing them to activate Mia OP without the lengthy tender process. We are also actively marketing this solution outside of Israel and initial customer responses in APAC and North America have been overwhelmingly positive.

Now to a more successful Gen AI-powered business line in the quarter, Voice AI Connect and Live Hub. Leading our revenue growth in the conversational AI business is the Voice AI Connect and Live Hub connectivity and orchestration services business, which grew north of 50% year-over-year. We delivered a standout quarter with strong performance across the board, highlighted by exceptional third quarter booking growth that puts us on track to exceed our full year target.

This momentum was fueled by high volume of new logo wins across the United States, Europe and APAC, along with significant expansion within our existing installed base. Driving this rapid growth is the emergence of the voice bots market, which is experiencing robust growth, driven by advancement in Gen AI and NLP. Market analysis projects that the voice bot market size will reach above $25 billion in 2034, up from just $4.3 billion in 2024, with a compound annual growth rate of 20%.

Now to the Voice AI Connect space. A key highlight was a high 6-figure voice, voice access project license agreement aligned with leading agentic platform, AI Agentic platform that's supporting virtual agent and agent assist use cases for its large enterprise clients. We view this initial engagement as the foundation for a strong and mutually beneficial partnership.

On the expansion front, we renewed a strategic agreement with a long-standing VoiceAI Connect customer, a leading multinational healthcare company. The expanded contract reflects a substantial increase in total value driven by the customer growing demand for virtual agent and assist capabilities as part of the digital transformation.

Additionally, we secured a significant follow-on order from one of the largest credit unions in the U.S., which is deploying our VoiceAI Connect solution for conversational IVR use case. Following the successful implementation of initial order in the first quarter of '25, focused on internal HR and help desk, the customer expanded rollout through its IVR this quarter, enabling self-service option for its end customers.

Moving on to Live Hub, offered as a Software-as-a Service. Live Hub is a cloud-native self-serve platform that helps voice bot developers for enterprise and service provider connect, connect, orchestrate and enrich the voice communication collaboration stride across various channels and systems.

During the third quarter, another exciting milestone was the introduction of Agentic AI capabilities within our Live Hub platform. This pivotal enhancement delivers an end-to-end solution carrying text-to-speech, speech-to-text and LLM-powered bot development with related best-in-class connectivity services, all tailored to service small to medium-sized customers. Importantly, our Live Hub financial momentum continues with ARR growing above 30% sequentially and substantially above 100% versus the year ago period.

Now turning -- before turning to the detailed business line discussion, let me quickly shift to the third quarter profitability metrics outlook. We performed -- outperformed on top line with revenue growing 2.2% year-over-year. Our non-GAAP gross margin for the quarter was 65.8%, which is above our previous quarter of 64.5%. The sequential improvement in our non-GAAP gross margin is attributed mainly to more favorable product mix and lower tariff related cost headwinds of about $0.5 million versus prior expense of above $1 million in the second quarter of 2025.

We expect fourth quarter '25 tariff costs to be in the similar range to this recent third quarter. Third quarter non-GAAP operating expenses of $34.7 million compared with $35 million in the second quarter and $32.5 million in the year ago quarter. On a year-on-year basis, the higher expenses are attributed mostly to targeted investment in growing the conversational AI business and higher impact from the weakening U.S. dollars against the euro in the third quarter.

Non-GAAP operating margin reached 9.5% compared to 7.2% in the previous quarter and 11.2% in the year ago quarter. Non-GAAP EBITDA margin was 11.2%, again, an improvement compared to 8.6% in the previous quarter. Non-GAAP earning per share was $0.17 compared to $0.14 in the previous quarter and $0.16 in the year ago quarter.

In terms of headcount, we ended the quarter with 961 employees, essentially flat across the first 3 quarters 2025 and compared to 935 employees in the year ago period. Net cash provided by operating activities was $4.1 million for the quarter and $25.2 million for the first 3 quarters of 2025. The key takeaway from these financial results is that our business remains strong and is expected to grow steadily through 2026 and beyond across our 2 primary sectors.

Looking ahead to the upcoming year, we expect a noticeable shift in our top line performance. Specifically, we project that 2025 will demonstrate both change and growth compared to 2024. This improvement is significant as it will mark a reversal of the declining annual revenue trend experienced in 2023 and '24. So, we're moving to positive trend, and we believe that '26 will be even higher.

Firstly, the UCaaS and CX connectivity business has stabilized compared to 2023 and '24. Additionally, 2 significant developments in the third quarter: one, signing the service agreement with the leading global system integrator and increasing engagement with Cisco, which is the second largest shareholder in the UCaaS market, involving all type of services, including Cloud Connect offering devices and more.

The 2 developments give us confidence that this connectivity business will perform well in coming years. And secondly, as anticipated, we expect strong growth exceeding 40% annually in our conversational business over the coming years.

Now to some of the major business lines, starting with Microsoft. Our third quarter Microsoft business was almost flat year-over-year, impacted by seasonality and late purchase order push into the fourth quarter. For the first 9 months of the year, Microsoft grew 4%, driven by our connectivity business, coupled with increasing attach rate of sales of devices, Voca CIC or Team Certified CCaaS and other conversational business application services.

Importantly, our pipeline of created opportunities remained robust in the third quarter, up 20% year-over-year and up 8% for the first 9 of 2025, again, compared to the year ago quarter. Market service and partner inputs continue to support a growth story for Teams Phone business, driven by the Microsoft Operator Connect program, where adoption in the market continues to show healthy growth.

Teams Phone usage is also strongly supported by Microsoft efforts to drive Copilot as a central capable chatbot for the Teams Phone meetings and calls. All this points to a strong market today and for coming years and further supports business expansion and dominance in the connectivity area.

Before wrapping up on Microsoft business discussion, let me share details of some representative wins. One is a very large greater than 1 million defense information system agency. Here, we have signed $1.1 million total contract value over the next 36 months through AT&T, covering the expansion of additional managed SBC services and calling plans in a new region.

Second win is with a financial services company operating internationally. It is a provider of investment management services outside to the U.S. This is a $1 million TCV contract over 36 months deals renewal of all prior services and purchased at a modest increase in value. Third is a win with a large -- one of the largest hospital, pediatric hospitals in the U.S., again, close to $1 million TCV over 36 months, covering live from managed services and gateways, enabling full migration from legacy PBX systems to Microsoft Teams.

Now turning to the contact center or customer experience market. CX grew by 13% year-over-year in the quarter, benefiting from growth in connectivity for CCaaS and connectivity services. We continue to see growing customer and partner interest in Live CX, which is an integral component of the live platform and targets applications such as cloud migration of contact center, replacing traditional 1800 services with click-to-call functionality and enabling conversational AI through Voice AI Connect and Live Hub connectivity.

As discussed in my earlier remarks, during the third quarter, we signed a landmark live platform agreement with a global Tier 1 system integrator where Live CX was a critical element of the broad-based agreement. Expanding our network of global Tier 1 integrator remains a key strategic initiative as it significantly broadens our addressable market. These partners focus on midsized customer experience customers, a segment traditionally underserved by our direct sales team. Importantly, our pipeline of opportunity remains -- remains robust and gives us confidence about our growth prospects for the balance of '25 and into 2026.

Now to conversational AI other lines. As discussed previously, conversational AI business grew 50% in the quarter. Key in the growth for the business line of Voice AI Connect and Live Hub, which we just discussed. Let's now discuss highlights of additional business lines that make up the conversational AI segment.

First, Voca CIC. We recorded another quarter -- record quarter of strong year-over-year invoicing and booking growth for Voca. Key highlights include major win in aviation. We signed a deal to deploy our team certified omnichannel contact center at a major APAC, Asia Pacific airport, one of the busiest airports in the world, beating out a couple of well-known premium CCaaS vendors. We won based on our ability to leverage our broad portfolio, offering a tightly integrated Teams-based phone and CCaaS service along with mobile app call enablement to contact center via our click-to-call solution.

Ongoing momentum in higher education, we continue to make solid progress in this vertical, adding another university this quarter that selected our best-in-class Teams certified contact center solution alongside our live Teams managed UC services. We now serve 12 university accounts in North America with Voca, including the second largest university in the U.S. and the largest school network on the East Coast.

Microsoft Unified certification, Voca became the second vendor worldwide to receive certification. We have distanced our self from competition as the only vendor with real-world enterprise production grade experience with this stack, thanks to our long-standing partnership with Microsoft.

Now to a new product update. Later in the fourth quarter, we plan to launch Agent Insights, which brings advanced conversational AI and generative AI to the Voca CAC platform. Powered by LLMs, it transforms recorded Teams interactions into structured insights, including AI summaries, sentiment analysis and one-click CRM updates. Each contact center desk can define customer summary prompts, ensuring precision and compliance across use cases.

Strategically, Agent Insights aligns with our unified integration model with Teams Phone, adding a critical AI layer to the Microsoft Teams CX ecosystem and strengthening Voca CIC as the role as the intelligent engagement layer driving efficient and quality business value. Now needless to say that Agent Insight is based on our technology developed in the meeting insight and therefore, we are in a good position to make great value and benefit from a technology in different areas.

Overall, our achievements are gaining recognition from leading industry analysts, culminating in a recent award from the UC today for best Microsoft Teams Contact Center, representing back-to-back win for the second year in a row in this category.

Moving to Meeting Insights. Meeting Insights Cloud Edition maintained strong momentum in this quarter with continued growth in new customer acquisitions. Other key metrics include the number of meetings and unique active users reached record levels, contributing to continued growth in monthly recurring revenue.

In addition to our broad market focus, we have developed workflow solution tailored to specific verticals, adding automation and connectivity to other leading enterprise IT solution aimed at leveraging Gen AI to enhance meeting productivity and accelerate business outcomes. Early traction has been promising.

One example involves the University of Central Florida, one of the largest universities in the U.S., which amongst a broad portfolio of solution customer takes from us, they deployed also Meeting Insight to generate AI-powered summaries and transcript of interaction between counselors and students.

Working closely with the customer, we perform analytics such as sentiment analysis and speaker [indiscernible] ratio, displaying key metrics in a custom dashboard available to the counselor, supervisors to support student wellness and improve graduation rates. This is just one example of how our vertical solution are transforming data into actionable insights and support workflows, optimizing outcomes. We look forward to sharing more in the coming future.

Moving on to Mia OP. Since second quarter, we have made significant strides in Israel and globally that are expected to drive growth in our conversational AI segment. In addition to the exciting contract award under Project Nimbus, we discussed earlier, our momentum in Israel is extending beyond the government vertical. We are now in final stages of several large tenders in other verticals such as healthcare and utilities, reflecting growing demand across various industries.

We also witnessed customer interest outside of Israel when customers understand the uniqueness of Mia OP solution that unlocks meeting intelligence at the edge computing level. Fresh from the debate of Mia OP in Asia Pacific in early third quarter, we are now engaging with several government opportunities in APAC countries in setting up a proof-of-concept trials. In late third quarter, we also showcased our solution in the United States and customer response was overwhelmingly positive.

We are currently in conversation with several U.S. federal and civilian agencies through a mix of collaboration with partners and direct engagements. We ended the third quarter with close to 10 customers in production and about 15 proof-of-concept project, all arising from word-of-mouth recommendation. Based on our exceptional pipeline of opportunities, we expect our momentum in Mia OP to further accelerate in fourth quarter and into 2026.

So, to wrap up our call, in third quarter '25, we continued to make solid progress in our long-term transformation to a hybrid cloud and voice services and Gen AI business application company. We delivered against our strategic objectives in that, a, we have a third consecutive quarter of revenue growth; b, we executed well to our playbook of leveraging our strong connectivity installed base in driving successful cross-sell value-add services. And third, the R&D and sales marketing investments we have made over the past several quarters have led to record conversational AI bookings in the quarter. And importantly, pipeline remains very healthy. This is the basis for our belief that we will grow in the next coming years more than 40% to 50% on an annual basis in the conversational AI business.

We are operating from a position of strength, supported by a fortress balance sheet, a dominant connectivity franchise and a growing conversational AI segment that enhances enterprise intelligence and productivity. We believe that these factors position us well for the rest of 2025 and increase growth in top line and earnings into 2026.

And with that, I have concluded my presentation, and I'll move over the call to the operator.

[Operator Instructions] We have a question from Joshua Reilly with Needham.

All right. Nice job on the quarter here. On the global Tier 1 system integrator win, maybe you could give us some more color on what helped you win that deal from a product perspective or any other factors that you think would be relevant to give to investors here.

Right. Well, I need to go back to the significance of our Live platform, which is a services delivery platform for UCaaS and CX. I think by now, this is the only platform that allows large system integrators, which serve large enterprises around the world, deliver all of the different services that are needed in order to move to modernizing the enterprise and to move to enhanced, I would say, communication and collaboration. Starting from connectivity, which connects all of the sites of a company across the globe. And then adding on top of that management, management of users, management of sites. And then on top of that, a list of business application and among them, an advanced and AI-first contact center, coding solution, meeting intelligence platform and now we're coming with voice bots and Gen AI applications.

So, all in all, this is the most advanced platform these days. And for a large system integrator that operates globally, this would be a great services delivery platform to serve its customers. And I think from that stems the recognition and the importance of that platform.

Got it. That's helpful. And then you're obviously building a lot of these kind of adjacent AI solutions for the communication landscape. If you look at the older products that you have in the market, whether it's FPCs or some of the gateways and all the older products that you sell, those are typically in pretty price-sensitive markets. What are you seeing with some of these new AI solutions and your ability to drive pricing power relative to the UCaaS market, which is historically a pretty price-sensitive market.

Right. Well, voice AI is a emerging market and therefore, those organization which are early adopters and quick to implement workflows and solutions that will substantially enhance their productivity are not less concerned with the cost. So, we do not see any price pressure at this point on the Voice AI business application. And we believe that as we will continue to enhance and add more features and make the solution substantially richer, we can still keep that. So, you have identified correctly the difference between the legacy business, which is price sensitive.

But again, there, we enjoy the fact that competition is becoming less and less powerful. But then we enjoy relatively convenient price environment, I would say, for Voice AI business application.

Got it. That's helpful. And then on the Microsoft business, I believe last quarter, it grew 6% year-over-year, and I think you said it was flat this quarter. Is there any change in the trends there? Or is that just really around the year-over-year comparison dynamics for the growth rate?

Right. So, I think overall UCaaS market is kind of flattening out in recent 12 months. We've seen that trend. It's been fairly strong up until '22, '23, then it becomes the expansion rate really decreased. It's a good market. It's a great market, right? Just take into account that out of -- if you go back to like 15 years ago and you talk about 400 million endpoints overall in the enterprise world served in the past by PBX'. So these days, UCaaS I believe, is serving less than $100 million. So, a lot of room to grow. And again, we all need to acknowledge that the majority of the growth occurred more in the U.S., U.K., Western Europe, Canada, Australia, maybe, et cetera. But there's a huge -- actually above 50% of the $400 million market that's still served by the old PBX technology. So, there's a lot of room to grow. So -- but pricing is such that I would assume that UCaaS will grow, but our services should be applied to the non-UCaaS market at a lower range. And I think that would be basically the driver for increased growth going forward.

Got it. And then last question for me is, if you look at the mix of revenue in the quarter, I would say that the product revenue was pretty strong, above what my estimate was and what I would have expected. Can you just help us understand maybe what outperformed on the product revenue side in the quarter?

Yes. As you've seen, first, we had a great quarter in terms of product recognized revenues. It was driven mainly at the software, which is part of the voice AI solution. So that's where the product growth came from.

As we have no further questions on the lines at this time, I'd like to turn the call back over to Mr. Adlersberg for any closing remarks.

Okay. Thank you, operator. I would like to thank everyone who attended our conference call today. With continued good business momentum in our UCaaS and CCaaS operations and continued growth in our emerging voice AI business, we believe we are on track to grow revenue and profitability in the next coming years. We look forward to your participation in our next quarterly conference calls. Thank you all. Have a nice day.

Thank you. Ladies and gentlemen, this does conclude today's call. You may disconnect your lines at this time, and we thank you for your participation.