Bentley Systems Aktie

Good morning, and thank you for joining Bentley Systems' Q2 2026 results. I'm Eric Boyer, Bentley's Investor Relations Officer. On the webcast today, we have Bentley Systems Executive Chair, Greg Bentley; Chief Executive Officer, Nicholas Cummins; and Chief Financial Officer, Bernard Andre.

This webcast includes forward-looking statements made as of August 6, 2026 regarding the future results of operations and financial position, business strategy and plans and objectives for future operations of entail Systems Inc. All such statements made in or contained during this webcast other than statements of historical fact are forward-looking statements.

This webcast will be available on Bentley Systems Investor Relations website at investors.bentley.com on August 6, 2026. And -- after our presentation, we'll conclude with Q&A.

And with that, let me introduce the Executive Chair of Bentley Systems, Greg Bentley.

Welcome, and thanks, as always, to each of you for your interest and attention. Bentley Systems positively exemplary operating results for '26 Q2 and the year as we see it, accord with our expectations to sustain foreseeably our dependable double-digit growth record. Early in 2025, I quantify that our outlook range for last year would serve to at least double over the 5 years following our 2020 IPO. The key financial metrics of ARR, revenues, profitability and SBC burdened free cash flow, and it can be confirmed that we met those sthesholds.

And now updating this and looking forward, a significant factor in my confidence is the boundless regeneration of demand within our infrastructure engineering end markets. with high economic returns globally and across sectors on investments in resilience, capacity and self-sufficiency. Within these priorities, relative proportions fluctuate presently most benefiting our offerings for integrated grid and for subsurface resources. But infrastructure engineering consumption has tended overall to remain predictably consistent, perhaps due to the constancy of engineering resource constraints.

To finally alleviate this engineering capacity bottleneck and thus further realize infrastructure investment potential through AI enablement is everyone's appropriate priority as the resulting benefits from improved infrastructure engineering, throughput and quality will be broadly shared across project delivery firms and infrastructure owner operators and all of us as their constituents.

Bentley Systems will especially benefit as our successive and multifaceted hybrid AI innovations continue for years to come on stream. I expect the established attended consumption of our software to continue growing as every day for every engineer will become increasingly valuable at the helm of ever more specialized AI leveraging applications.

But this business model will, in due course, be supplemented and potentially multiplied by monetizing agentic API consumption of our modeling and simulation software primarily to optimize designs at machine speed, accelerating the proprietary AI strategies of infrastructure engineering firm and our emerging asset analytics offerings monetized through asset consumption subscriptions per asset are breaking through to finally leverage digital twins and operations and maintenance for infrastructure owner operators.

Adding to my confidence in looking forward, these incremental AI-led consumption opportunities are uniquely accessible to Bentley Systems by virtue of our established positioning as the major comprehensive infrastructure engineering software provider across both the leading infrastructure project delivery and owner-operator enterprises.

We have earned our standing as a trusted digital Quartermaster for each of these major infrastructure engineering organizations over decades of proactively embracing and incorporating potentially disruptive technologies and business models so that no 1 ever needs to start over in order to stay ahead of the innovation curve.

In my now long experience, the winning strategy for every technology opportunity is hybrid, integrating, in this case, AI advances within accretive overall continuity. Facilitating progressive enterprise AI adoption is the invigorating current priority for our 1,000 success force engineers embedded in E365 accounts.

Last quarter, I talked about AI's auspicious economic leverage for engineering firms whose work pre-AI has been constrained by the limiting supply of infrastructure engineers. Underscoring our incumbency advantage in helping these firms to leverage their engineers with AI agentic assistance to perform more work and particularly to optimize designs, 470 of the 610 engineering news record global top design firms ex-China, our BSI accounts, averaging ARR of nearly $1 million each.

I will now similarly quantify our comparable point of departure for the leading infrastructure owner operators. For decades, the authoritative global ranking of the largest owner-operator organizations measured by their fixed tangible asset value net of depreciation, has been the annual Bentley Infrastructure 500 top owners rankings. The upcoming 2026 BI 500 will be published on bentley.com.

The most recent 2025 BI 500, it no longer includes Russia, own and manage about $21 trillion of net infrastructure assets. Not quite half of those top owners assets are in the public works/utilities infrastructure sector, followed by industrial and then resources sectors, which together comprise most of the balance. In net asset proportions, the top owners are located about half in the Americas, followed by EMEA and Asia Pacific.

In particular, the 43 top owners in China account for just under 10% of these assets and because it's with Chinese state owners that we encounter unique geopolitical obstacles currently, to quantify BSI penetration, the following charts are ex-China. Over 3/4 of the ex-China top owners, managing Mel over 80% of such infrastructure assets or BSI accounts.

Excluding top owners in the commercial/facilities sector, where we are less vocal, 90% of ex-China top owners infrastructure assets are managed by BSY accounts. 153 ex-China top owners, holding the majority of these accounts net infrastructure assets have already adopted Bentley Infrastructure Cloud with most using project-wise. They are thus accumulating engineering data for delivered projects, which will make AI-enabled digital twins more valuable.

And quantifying the BSY spending by these top owners I use current year run rate, which beyond ARR includes our relatively minor amounts of license sales, professional services and other subscriptions, to fully capture the offerings which are exclusive to owner-operators of cohesive and for acid analytics.

These 346 top owner accounts spent annually over $330 million in run rate with us, averaging about $1 million per year for those not in the commercial/facilities sector, and collectively representing about 20% of our overall business. Accordingly, annual BSY expenditures currently averaged $21 per million of the $15.5 trillion of net infrastructure assets owned by these 346 ex-China top owners or our BSY accounts.

Introducing and fully developing infrastructure AI, it is a tremendous and literally unique advantage for Bentley Systems to be in the pole position with virtually all of the infrastructure engineering organizations with the most to gain in both project delivery and operations and maintenance. For every top design firm and top owner account infrastructure engineering and enhanced Bentley systems are mission-critical factors of their production.

It is certain that in each case, their current expenditure levels on software in proportion to their respective engineering labor costs and asset values will be multiplied by orders of magnitude as AI is inexorably and advantageously integrated to improve infrastructure, capacity, quality and economics.

In sum, I believe this enterprise account springboard will continue foreseeably to improve Bentley Systems on economics and growth prospects. So at this point in time, when investable sectors seem ever more subject to comparisons from first principles, -- let's update our own point of departure. Free cash flow as a valuation benchmark has the advantage of meaningful commonality across investment sectors.

However, it's uniquely salient for software, in particular, given the opacity of 606 subscription accounting other than for BSY, with our consumption dominated revenues being recognized primarily ratably to the virtual exclusion of multiyear noise. Here has shown in the past 5 years of BSY's free cash flows aggregated within trailing 4 quarters ending in each Q2 and for comparison to the latest for '26 Q2.

Over this period, these annual cash flows compounded at an average rate of 14.7% annually to reach $498 million over the last 12 months. For mature software companies, another rightful valuation consideration is stock-based compensation, given its prevalence and typical magnitude, I don't consider that cash flow should be counted as free to the extent that it needs to be expanded for stock repurchases to offset resulting dilution.

Shown here as accordingly offsetting free cash flow is BSY's operating, that is not acquisition-related, stock-based compensation for these periods. The remainder is BSY's truly free cash flow as burdened by operating SPC. So burdened free cash flow reached $426 million over the last 12 months, having expanded over this period at a CAGR of 16.7%.

Our consistent low double-digit annual growth rate in ARR compounded by our established annual improvement of about 100 basis points in AOI less operating SBC margin served indeed to double over the last 4 years, this valuation metric, which to me seems most economically appropriate to us shareholders.

And consistent with Bentley Systems conscientious stewardship of stock-based compensation, over most of our public history, we have tended to allocate free cash flow to stock repurchasing an approximate keeping with annual requirements to offset SBC dilution. Here are the quarterly expenditure amounts for all repurchases, including de facto repurchases associated with net distributions through this period up until '25 Q3.

Even though BSI has had a discretionary share repurchasing authorization throughout, there hasn't been a very significant opportunistic that is inverse correlation between our share price and overall repurchase expenditures. What changed since late last year is -- by then, we had resolutely worked down the debt leverage we incurred to finance the platform acquisitions of Sequent and Power line systems in 2021 and 2022 to a tolerably optimum range of about 2x.

This has enabled us, since then to allocate more capital to discretionary stock repurchases without changing either on ongoing cash flow funding for programmatic acquisitions nor on balance sheet preparedness for potential larger scaled platform acquisitions. What eventuated during '26 Q2 was a stock price, which at our marginal financing cost enabled repurchasing to be accretive just on financial fundamentals to a significant and reasonable basis for valuation.

We variously repurchased 3.1 million shares during Q2 and subject to remaining within an optimal leverage range, I expect us to continue to responsibly act upon any such opportunities going forward. Indeed, the net result of this SBC and corresponding repurchasing discipline over this period has been to avoid share dilution.

Shaded here at top is the portion of our fully diluted shares attributable to our convertible debt. In fact, the redemption of our 2026 mature and convertible debt during '26 Q1 and reduced our fully diluted share count by about 3% as will presumably recur in '27 Q3 with the maturity of our remaining convertible debt. So reflecting a compounded average growth rate of negative 1% through this period. At the end of '26 Q2, our fully diluted share count was down to 319 million shares.

And we thank you for being among or interested in coming or informing those of us who are the holders of those shares. And now over to Nicolas, and then Werner to cover this quarter's development. Thank you.

Thank you, Greg. We had another strong quarter, executing consistently against a backdrop that keeps working in our favor. The world needs more infrastructure and resources, and it needs them faster than they can be delivered. Everywhere we look across accounts, large and small, the constraint is the same. There are not enough engineers which is why engineering productivity, making every engineer both more efficient and more effective is at the core of our AI strategy.

Let me pick up the AI threat from last quarter. At the end of 2025, we launched our infrastructure AI initiative. And last quarter, I reported that leading engineering firms and owner operators were asking us to instrument our applications to power their own AI-driven workflows. I also shared that we had released our first MCP server for Stan.

This quarter, I want to show you how far we have come and why we are confident in the approach. Our conviction is that when it comes to mission-critical infrastructure engineering or applications and today's AI models are far more powerful together than apart because each does something, the other cannot. Our applications are deterministic. They perform the engineering itself, the modeling, the analysis and the simulation, and that work is trusted because it has been proven over decades embedded in workflows across infrastructure value chains.

AI models are different in nature. They are probabilistic. What they contribute is natural language processing, high-level reasoning, and the ability to break a problem down and generate instructions that our applications then execute with engineering precision. The MCP server is the interface between the 2 turning the AI's instruction into real validated work inside the application.

One point I want to stress, we are deliberately open. This is not a walled garden. Our accounts can pay our applications with which every assistant and whichever model they have to standardize on, whether Benicopilots, Anthrapic Claude, Google Gemini or open AP -- our aim is to be the trusted engineering layer beneath all of them, whichever AI model happens to lead at a given time.

Now to the progress. Last quarter, we had released a first NCP server for STAD. Since then, we have released 5 more across bandy open applications with more to come. The response from our accounts has been very positive, once aggress what is possible. This is still early and a great deal of our work today is education on 2 fronts.

First, we are helping users cut through the considerable noise around AI and we have made this a priority with several campaigns underway to show what is generally achievable now. Second, we're staying close to our accounts as they adopt. And here, application engineers and solution architects are proving invaluable serving in effect as 4 depot engineers, helping our accounts evaluate and integrate these new capabilities into their workflows.

Finally, on the commercial model. As discussed last quarter, our next steps were to instrument more applications and to validate the commercial model for this new usage pattern. The first is well underway. On the second, our priority remains in order adoption, exploration and validation with monetization to follow, and we continue to be transparent with our accounts about that sequence.

Earning adoption and trust first is precisely what will let us capture our fair share of the value that will be created as our applications are used at machine speed with AI.

Now turning to our business highlights. Our year-over-year ARR growth for Q2 accelerated to 12%. Our net revenue retention rate remained high at [ 19%, ] consistent with previous quarters and underscoring the stability and growth within our existing accounts. Our enterprise fed commercial program continues to drive steady growth. We were particularly pleased with our renewals in the quarter as Q2 is typically our second largest quarter for renewals.

This strong performance gives us confidence for the rest of the year. New logos contributed again 300 basis points of ARR growth primarily within the SMB segment. Through Virtuoso, our flagship commercial program for SMB accounts, we again added over 600 new logos in Q2. The underlying SMB market sentiment remains positive with accounts reporting healthy project backlogs extending well into 2027.

Accounts are increasingly viewing bended technology as part of their business backbone rather than point solutions for specific projects, which validates our cross-selling and upselling efforts.

Turning to our performance by infrastructure sector. Resources was our fastest-growing sector in total driven by mining once again, with strength across geographic regions. The fundamental drivers for this demand are macroeconomic and long-term. Countries around the world are increasingly prioritizing sales efficiency given ongoing geopolitical tensions and supply chain disruptions.

In addition, the global push for electrification, including to power AI data centers also depends on securing critical minerals. Our largest sector, Public Works and Utilities delivered another strong quarter. driven by sustained infrastructure investment worldwide. Within Public Works & Utilities, Power Line Systems, or PLS, continues to be the primary driver of growth in our electric grid business benefiting from strong global demand for grid transmission, capacity expansion and resiliency.

I will take a deeper dive into PLS and our broader electric grid portfolio shortly. Growth in industrial sector continued to be solid, while commercial facilities remain relatively flat.

Turning to our tone of business by geographic region. In the Americas, our largest region, the U.S. continued to deliver strong growth. The underlying fundamentals of our accounts remain very strong, characterized by double-digit backlogs and a bullish outlook on their long-term growth, driven primarily by transportation water, power and data centers. Public funding at the federal and state level remains robust, supplemented by a healthy influx of private capital funding.

Latin America delivered another very strong quarter, led by mining and an increased focus on transportation in the region. EMEA delivered a solid quarter as the overall region remains well invested. The quarter benefited from strong renewals at a number of large accounts. Large ongoing natural infrastructure programs are driving demand in the U.K. Fundamentals in Europe are also strong.

In Germany, while the EUR 500 billion infrastructure fund is in place, actual deployment has been slow as early firms are backfilling existing deficits, delaying the impact on new projects. In the Middle East, despite the ongoing conflict, accounts have returned to work, consumption has rebounded and deals are progressing again. Asia Pacific delivered strong growth with Australia leading the way as performance bounced back strongly followed closely by India.

China representing only about 2% of ARR, continues to operate against the same economic and geopolitical headwinds. Across most of the region, Rail is a massive long-term opportunity with major projects in Australia, India, The Philippines and in Indonesia. Offshore oil and gas is also an investment priority across the region, driven by major field developments in Southeast Asia and offshore redevelopment in India.

We are well positioned for this work with our Saks offshore design and analysis engine.

Now I would like to highlight the continued outperformance of our power line systems. PLS is the goal standard for the design and analysis of overhead electric power transmission and distribution lines and their structures. It is part of our broader electric grade portfolio, encompassing open utilities, substation plus and SPDA that together allows us to address the full spectrum of great infrastructure from transformation and distribution to substations.

Looking back 4.5 years since the acquisition, PLS has become a vital part of our core business and the financial pillar of our electric grid offering. We have also seen significant growth internationally, leveraging Bentley's established global reach and go-to-market engine. In fact, PLS revenue outside of the U.S. is now as large as the entire PLS business was when we acquired it in 2022.

Reliable energy delivery is more critical now than ever as electrification and the rapid expansion of AI data centers placed unprecedented stress on the global grid. In the U.S. alone, the network requires an estimated 35 gigawatts of additional capacity by 2030. PLS software is instrumental in bridging this power gap. We're seeing our users apply PLS at impressive scale to overcome real-world physical and operational challenges.

For instance, Exodigitally stress-tested century old transmission towers spanning the Ohio River, allowing the TT to engineer targeted stabilization that saved $80 million and avoided up to a decade of permanent delays. And when 120 miles an hour winds destroyed a transmission quarter in ToF and associates use Petascad to redesign the network digitally accelerating the rebuild to restore power 18 days ahead of schedule. The vibrancy of this ecosystem was on full display during Q2 at our biannual PLS User Group Conference in Madison, Wisconsin which drew more than 500 attendees from nearly 300 companies, including over 90 utilities representing 20 countries.

A central theme of the conference was AI where we announced 3 new NCB servers for PLS products, including PlasGrid. This is an important milestone because it demonstrates how with MCP servers, we can achire AI systems to not only interact with engineering applications, but also directly with rich engineering data by connecting AI Superior's great via MCP users can query vast digital twin repositories in natural language, ask you questions like find the 100-week structures in my entire grid or which transmission line spans will thermally limit the interconnection of the proposed AI data center.

Unlocking this engineering data directly is precisely the direction we're taking across a broader bed infrastructure cloud enabling our users to extract actionable intelligence across their project files and asset information at machine speed. With PLS as a critical component of our electrical grid offering and very much part of our core now, Bentley is uniquely positioned to continue to benefit from the massive investments required to power an electrified future.

In summary, Q2 was a strong quarter and we enter the second half of the year with great confidence in our disciplined execution and market fundamentals. And with that, over to you, Werner, for a detailed review of our financial results.

Thank you, Nicolas. We are pleased with our performance for the first half of the year. Our second quarter results extend the momentum from the first quarter and set us up well for the balance of 2026. Total revenues for the second quarter were $411 million growing 12.8% year-over-year and 12.2% in constant currency. For the first half of the year, total revenues increased 13.6% and or 12.1% in constant currency. Our performance continues to be led by our mainstay subscription revenues, which represented 92% of total revenues during the quarter.

Subscription revenues increased 13.6% year-over-year or 13% in constant currency, reflecting continued strength across both our E365 and SMB initiatives. For the first half of the year, subscription revenues increased 14.1% or 12.6% in constant currency. In our small and less predictable revenue streams, Services revenues increased 9.4% or 8.7% in constant currency, driven by continued improvements in maximum related services activities within cohesive extending the recovery trend we discussed last quarter.

For the first half of the year, services revenues increased 18.8% or 16.5% in constant currency. Perpetual license revenues for the quarter were approximately $10 million, down roughly $0.5 million year-over-year and approximately $19 million for the first half, down about $2 million. Perpetual license sales remain a very small part of our business with approximately 2% of total revenues.

As of June 30, our last 12 months recurring revenues were $1.486 billion, an increase of 13.5% year-over-year or 11.8% in constant currency and represented 93% of total revenues. Our last 12 months constant currency account retention rate remained consistent at 99%, and our constant currency net revenue retention rate remained at 109%, consistent with recent quarters. The combination of our high retention rates and new business momentum gives us confidence in the continued durability of our recurring revenue growth.

Now turning to ARR. We ended the second quarter with ARR of $1.536 billion at quarter end spot rates. On a constant currency basis, our year-over-year ARR growth rate was 12%, and our sequential quarterly growth was 2.9%, all organic and in line with our expectations for the quarter. We continue to expect our quarter-over-quarter ARR growth seasonality to be similar to 2025 and thus organic year-over-year ARR growth rates to be relatively stable during the year.

Our GAAP operating income was $89 million for the second quarter and $215 million for the first half. As I've discussed previously, our GAAP results can be impacted by deferred compensation plan revaluations and other acquisition-related items. Moving to our primary profitability measure, adjusted operating income less operating stock-based compensation, or AOI less operating SBC, AOI less operating SBC was $116 million for the quarter, with a margin of 28.3% and $257 million for the first half with a margin of 30.8%.

This performance was in line with our expectations with our first half margin trailing the prior year as we weighted operating investments early in the year compared to 2025. Our first half G&A also reflects our new enterprise-wide finance and Quote-to-Cash platforms, which went live during the second quarter, driving a step-up in cost that won't go live that was contemplated in our outlook.

Notably, we did not adjust these costs out of our profitability metric. We absorb them within our margin commitment underscoring the quality of our first half performance, while positioning us for greater efficiency and scale. We remain well positioned to deliver on our annual constant currency margin improvement. Our free cash flow for the quarter was $64 million and $252 million for the first half.

This result was in line with our expectations and reflects 2 key factors we signaled on our last earnings call. First, our 2025 free cash flow benefited from exceptionally strong collections at year-end, which, as anticipated, created a tougher year-over-year comparison in the first half most pronounced in the first quarter. Second, our plan to weigh operating expenses more towards the first half this year is reflected in our year-over-year comparison for both profitability and cash flows.

As a result, and consistent with the framework we shared in our outlook, first half free cash flows represented approximately 47% and of our full year outlook, in line with the 45% to 50% range we guided for the first half. Looking beyond quarterly timing, on the last 12 months basis, free cash flow of $498 million was up 15%, and we remain on track to meet our full year free cash flow outlook of $500 million to $570 million. We continue to execute a disciplined and balanced approach to capital allocation.

During the quarter, we closed on a new $550 million term loan A under the Calian feature of our credit facility. This transaction was completed at attractive terms and used to repay outstanding borrowings under our revolver, lowering our interest costs. This provides ample capacity to support our strategic priorities, including addressing our mid-2027 convertible notes maturity while also funding potential programmatic acquisitions, ongoing share repurchases and dividends.

During the first half of the year, we reduced net debt by $32 million and returned capital to shareholders by deploying $155 million for share repurchases, up meaningfully from the prior year and $42 million for dividends. Our balance sheet supports significant strategic flexibility. At quarter end, capacity under our credit facility was $1.2 billion, and our net debt leverage was 1.9x adjusted EBITDA, consistent with the prior quarter.

We continue to actively manage our interest rate exposure, our safeguards include the low fixed coupon on our remaining convertible notes and our $200 million interest rate swap expiring in 2030. Overall, our performance through the first half of the year compared favorably with our expectations. We delivered consistent growth in revenues, recurring revenues and ARR and while maintaining disciplined profitability and cash flow generation to remain comfortably within our full year financial outlook. This year, that outlook also includes a range for AOI less operating SBC, reflecting the annual constant currency margin improvement I referenced earlier.

With regards to foreign exchange rates, through the first half, the U.S. dollar strengthened relative to the exchange rates assumed in our 2026 annual financial outlook resulting in approximately $5 million less revenue from currency, approximately $2 million in the first quarter and $3 million in the second. If end of July exchange rates were to prevail for the remainder of the year, our second half revenues would be negatively impacted by an incremental $8 million to $10 million relative to the exchange rates assumed in our 2026 outlook.

And with that, over to Eric for Q&A. Thank you.

[Operator Instructions] And with that, our first question comes from Matt Hedberg from RBC Capital Markets. Thanks, Bernard. Before we begin, I just wanted to remind everyone just limit themselves to 1 question. And with that, -- our first question comes from Matt Hedberg from RBC Capital Markets.

Sorry, sorry about that. Can you hear me okay now?

Yes.

Excellent. Thanks for the question. Congrats on the results. And the ARR growth at 12% was certainly impressive here, the acceleration versus last quarter. I guess what are the keys now as we get to the back half of the year on sustaining that growth or even improving?

And I know we've talked about maybe getting to the high end or above that and 12%, you're close to that. But I guess I'm wondering like continued success from asset analytics? Is there a macro element, anything -- anything that could kind of help us on that trajectory.

Yes. Thank you, Matt. So yes, definitely, we were quite pleased with the performance consistent with Q1, an acceleration, nevertheless, because of momentum in resources, which is still our fastest-growing sector and in particular, in mining. And then just general strength in public works utilities, which includes electric grid, but we've seen strength across the sector. So the momentum is definitely there for us to get to the upper part of the range, we will need this momentum to continue, and there's no sign why it will slow down.

But we will also need to do an acquisition potentially -- and yes, bringing some big deals with asset analytics, which is a rather lumpy business, as we discussed in previous calls, right, where we depend a lot on big deals.

So I'd say each of those factors that Nicholas mentioned are relatively likely, but they all have to happen together to wind up at the top end of the range and until we have.

The next question comes from Joe Vruwink from Robert W. Baird.

Great. the discussion on your older customer base, I thought it was interesting, and I know it's not the same basis for comparison that you highlighted last quarter with project delivery firms and kind of your earn rates with those customers. But just the $21 in earn rate for Bentley relative to the $1 million in assets. Where do you think that can go?

And I guess inherent in the question is you've grown your portfolio of solutions that are relevant for owners quite a bit over the last 15 years. And so do some of the recently acquired IP or just the way that infrastructure cloud is evolving, unlock a much bigger opportunity with that side of your customer base?

Well, it's all of the opportunity with infrastructure and AI to improve the quality, resilience, the life cycle of the infrastructure assets, the other operators benefit from better CapEx, which optimized designs will provide in terms of constructability, less subsurface risk and so forth, things we're working on through our open applications. But just as you say, the biggest opportunity is in operations and maintenance and especially for optimizing that to do only what's needed and what works and for AI to help discern that by operating on badly infrastructure cloud and using the as operated inspection and monitoring that's made possible through asset analytics.

So we think we've closed that circle conceptually, leaving a lot of opportunity, and we should monitor that $21 million of net assets just as we do on the correspondingly on the engineering firm side to see how they're going to go about improving their economics by spending more on software and AI together, as Nicolas described.

The next question comes from Jason Celino from KeyBanc.

Great. Thank you for the update on PLS. I think it's a business that we often kind of forget about -- I think when you originally acquired that asset, it was kind of growing similar to Bentley corporate average, but its margins were meaningfully better. And then I think entering the year, when we thought about permitting reform, that was 1 of the businesses that could really benefit from it.

Are you able to maybe provide an update on maybe the financial profile of what PLS looks like today? And if you've seen any acceleration in the business from any type of permitting reform?

Actually, PRS has been a growth engine from -- almost from the time of the acquisition and very consistent. The growth has been both in the U.S. and internationally. You heard in the prepared remarks that now our business with PLS outside of the U.S. is as big as the entire PS business was when we did the acquisition. So we grew very well outside of the U.S. But in the U.S. as well.

And in the U.S., we continue to grow despite the lack of permitting reform because there's a lot of investments going into the existing grid, just to make sure that it is capable of keeping up with the demand for more electricity that it is resilient in the face of extreme weather events, et cetera. Therefore, we are well positioned to see our PLS ARR even accelerating as permitting reform goes through. Permitting reform, there's always a lot of discussion. It is, in a sense, a bipartisan topic. Everybody wants it. But we need Congress to get its act together, no pun intended.

And conversations are resumed about permitting reform. And a fantastic vehicle for permitting reform would be the new surface transportation bill which has been passed by the House of Representatives, but is now a bit on hold until a bigger decision is made. But yes, there is just a lot of activity for permitting reform in the U.S. And when -- if and when I think it's more about when this goes through, then this will help further grow our P&S business, which is already doing very well.

Next question comes from Daniel Jester from BMO.

Sorry, camera for some reason it's not working. So Greg, in the last couple of quarters, you've talked about sort of the accelerating of the art of what's possible in AI. And in the prepared remarks today, you commented about sort of the opportunity and what you're seeing in your customers. I guess new MCP servers, you've got a lot of opportunity here. I guess from a customer enablement perspective and their ability to harness these tools, are...

I think Daniel was going to ask where do they stand? And that is -- my point was we're embedded closely with the largest engineering firms and operators who are especially the engineering firms in a hurry because of the resource capacity constraints, they could do more business if they could improve efficiency and effectiveness, as Nicolas said, -- so they're in a hurry. But they understand the best way to get there is a hybrid approach where their AI assistance, their own agents would take advantage of established functionality.

And our education is how to put those 2 together into a hybrid approach. I mentioned hybrid, particularly as something which over my career, I've seen repeatedly be the result of innovation way because there isn't something that's going to take over 1 particular phenomenon, but a combination together. In addition to the nature of the hybrid, Nicholas described of inference on the part of assistance calling established deterministic engineering logic.

Another aspect of hybrid that I expect to come about is in the computing form factor, which will include some, a lot of or choices for what we could call sovereign AI computing on local and edge devices because in the case of engineering firms of the sensitivity of their intellectual property in the case of owner operators of the cyber risks involved. So in rate, their hybrid directions are going to be where things settle out in the long term. And you probably were asking me about the long term given my longevity.

Next question comes from CD Panigrahi from Mizuho.

Just to continue to the prior questions, the annual question, maybe I'll extend that. I mean the opportunity you talked about AI, like, can you give us more concrete signal even directionally, like whether this AI or the commercial model, API consumption, all this. When do you think is kind of going to -- like you can monetize? Is it more end of '26 or '27 even kind of thing.

And in that context, you talked about the data and Interface, whether cloud or Charity, does it help Bentley system as a platform? Or does it create opportunity for stand-alone companies to pipe the data into these interfaces. How do you differentiate there when Claude charge real become an integration layer?

All right. So let's go after these different questions 1 by one. So the signals we're getting from the accounts we're engaging with on this novel way of using our engine applications. The signal is very positive. The effort on our side is really to cut through the noise because there's so much noise around Obviously, every software provider under the sun is approaching the different accounts, talking about their own abilities. So we need to cut through the noise.

And then when infrastructure organizations realize what we're talking about and the potential there then the reaction is very positive. We hear words like, oh, this is a game changer. This is changing everything. There's a lot of value that's going to be created and so on and so forth. So that's very encouraging. And therefore, we are really sticking to the phasing we talked about before, which is our #1 priority is adoption, exploration validation, right, getting -- making sure that accounts are aware of those capabilities, they're really adopting those capabilities, and they are validating the potential value there and monetization next.

And we're very transparent about that exact sequence with our accounts saying, at some point, of course, we will need to monetize. And the reaction from the accounts on the monetization is a shared understanding that indeed the trail way for us to monetize, which is attended consumption with 365 or user-based subscriptions, and it's all about users wouldn't quite work going forward when so much value is going to be created with AI itself interacting with our applications and using them at machine speed. And therefore, the metrics have to change.

So we're taking this time of adoption, exploration, validation of our applications at the time also to validate what is going to be the potential commercial model. So that when we start to monetize, there is great receptivity our accounts on the way we're going to monetize that. We're not planning to monetize this year. We're planning to start monetizing next year.

And you should expect us to monetize in a very reasonable way. What we don't want is a big price tag to become so an obstacle for infrastructure organization to use these capabilities. It will be really going against what we're trying to achieve here. We want to unlock a lot of value, and we want to lock a lot of value for all the players of the value chain from the end clients because they will end up having much better design because they've been optimized at machine speed for the engineering firms themselves because potentially this is the opportunity for them now to evolve also their commercial model and capture a fair share of the value that's going to be created and for us as well as provider, right? So we've been very thoughtful, very diligent on how we're introducing these capacities.

And then to your last question about us having this very open approach and letting our users use whatever AI systems they want and using those assistants to interact with engineering data that is coming from our systems or third-party systems. We actually want that. We welcome that, right? What we've learned with all the infrastructure organization we've been engaging with in the context of the infrastructure AI initiative is that, we need to be ready for all sorts of permutations and how exactly our applications are going to be used, how beneficial is going to be used, how data is going to be used, right?

Some of them are saying we want to use the Bentley AI system called Bentley copilot because it's a much better user experience, the users stay with the same applications. Others are saying, no, no, actually, we want to continue to use in other AI systems, for example, Claude because we've been tailoring it with data that is specific to us as an engineering firm, for example, and we don't want users to have to swap and go from 1 AI system to another. We want them to continue to use exactly the same.

That's the kind of feedback that we're getting, which is a lot of creativity, and we need to be ready for all sorts of use cases. In that context, we're very clear that where we are bringing value, where we're adding value and ultimately where we need to monetize is with the underlying engineering applications. And then with the -- and benefactor Cloud overall, regardless exactly how these applications are used, whether it's with our own co-pilot or with third-party assistance.

Regardless our data, the data that has been created with our platform is being used, whether it's through our co-pilot world with third-party AI assistant. We need to be completely open to that, and not having any artificial limits in order to do this. Again, we will monetize at the underlying level at the underlying level, which the engineering applications themselves. -- and then infrastructure cloud.

The next question comes from Kristen Edman from Oppenheimer.

Wanted to dig into the sequence results a bit and specifically in mining, we're seeing a lot of capital discipline in the space -- so I'm wondering if you can unpack for us how much of the growth are you seeing from greenfield versus maybe brownfield expansion versus just this general trend toward adopting digital tools.

And then somewhat related, I'm also seeing across my coverage, a lot of M&A in the digital mining technology space. So when you look at your acquisition pipeline, is there -- is this an area where you'd see opportunity for inorganic growth?

Yes. On the -- thank you, Kristian. On the first point, so yes, 2026 is probably going to be a record year in terms of investments in CapEx in mining with -- and we see it around the world, across geographies. So it's very is broad-based. And the kind of investment we're seeing is primarily brownfield in order to have a much faster return on investment than high risk, very long-term return on investments, greenfield investments. So it's primarily brownfield.

We're seeing it also with the use of our software as we're tracking it, that it's primarily used for brownfield rather than greenfield. And why is that? It's because of this underlying trend of sales efficiency countries around the world who want to get to critical minerals on their own and fast because of the geopolitical tensions, because of the disruption to the supply chains, I think there's a wide understanding around the world on the critical nature of those minerals and they need to be able to access them without obstacles. That's the underlying trend.

Therefore, our growth as well with our software, which is being used for exploration and operations of mines. It's also very broad-based, and we've also seen growth around the world. We've seen it with large accounts and smaller accounts especially with the midsized mining company. It's very encouraging because it's so broad based, it means this is quite durable from that standpoint.

Then in terms of M&A, our appetite remains the same. We did indicate already last quarter that besides asset lytics, which remains a big priority. We are definitely looking for other opportunities, including in resources. If we see some assets out there that can help us exceed our vision that can help us excel our strategy and they can fill some gaps that we have in helping mining companies get a better understanding of the subsurface, then we'll do it. Whether by rates for mining, it could be also in geothermal, other sources of energy, it right resources overall.

It's still our expectation to be able to do that this year.

The next question comes from Faith Bruner from William Blair.

Maybe just wanted to jump into some of the AI noise. It seems like lately, we hear a new announcement from anyone talking about some new AI solution, whether it's a broad infrastructure use more specialized. So maybe what are you guys thinking about as the competitive landscape kind of shifts? And how can this maybe play into your open ecosystem approach for AI?

There's a lot of noise in our space, maybe not so much in core infrastructure. We're going to see it in adjacent spaces like AC or let's say, buildings. There's quite a bit of activities there. They're indicative of where things could go. What's quite clear is that a lot of use cases are more at the edges of what we're doing. When it comes to using AI in, let's say, core engineering and engineering for infrastructure, there's really no better value proposition than this combination of our own engine applications that are trusted, that are already integrated in workflows across infrastructure value chains around the world, together with third-party AI assistance, AI models, L&Ms, right?

And that combination is here and now. So this is not a start-up activity. It's not a -- this is where it could go. It's already there. Those cables are there. And so when we're engaging with our accounts, this is where the conversation is, which is let's not go into what very forward-looking conversations. Let's just talk about what is possible right now, and this is where typically the eyes and opened very, very wide when they realize the capacities, the possibilities, the value that can be created today already.

I myself am glad to hear each occasion of engineering organization, infrastructure engineering organizations adopting AI for whatever it is they're adopting it for. And in my ideal, these organizations would take pride in promoting and marketing their proprietary AI approaches to their overall strategy to optimize the sun. And heuristics you can imagine learning applications and so forth, that would be particular and specific and proprietary.

And there have been in the past waves of innovation in engineering and design that where organizations competed on the quality and differentiation of their approaches. In all cases, that can be accelerated if they use existing functionality for the modeling and simulation so that they don't need to take on building up from the ground, what exists already.

It's in the -- how the solution space is explored and learned from and what you optimize and so forth, that there is so much to be gained and that should be proprietary and specialized and differentiated and AI can accelerate that for our user organizations. And that's why the open approach that Nicholas described is the right month for us.

The next question comes from Jay Vicor from Griffin Securities.

The third part of the question is you for the quarter and for the year-to-date, your increase in sales and marketing was fairly considerable and more than the increase in R&D on a small base spending. What is your expectation for R&D for the remainder of the year and into particularly when we think about the recent uptrend for example, in your engineering open positions, which have been noticeably higher over the last few months.

And then the technical side of that question is for Greg and Nicolas, could you remind us what your R&D and developmental priorities are, including but not limited to your design side of the portfolio, which you've been highlighting as a priority since 2022, but maybe bring us up to date on all that.

Well, maybe I'll start with the latter, JV because then it's a good context for the conversation about where exactly we're investing. So the priorities are AI invention applications, AI with Penn Infrastructure Cloud and then synergies across the full portfolio, including our Sequenotfolio. So across, if you want be infrastructure cloud and Sequent. Those are the priorities from an R&D standpoint.

And then within AI fringing applications, then it's both building our own capalities, which we haven't talked about this call. but also instrumenting applications to interact with third-party AI, which we've discussed at length now, right? So those are the priorities.

So when it comes to investments, I will say going forward, we're leaning very hard in adopting AI internally and across all functions. What I think we can expect that in the longer run, we grow certain functions as a percentage of revenue faster than others. And you may have heard some of us talk about growing R&D in particular. And of course, we will do that as a software company. There's so much potential for AI that we want more engineers, and we want them to work much faster, thanks to the AI capacities.

But most probably will also be an increase of spend as a percentage of revenue into go-to-market functions because of the point I mentioned earlier, which is there's so much noise around AI that we need to cut through it. And that does require investments in marketing, and it requires investments also with our success force, application engineers, solution architects to be right there as for deployed engineers with our accounts to make sure they're aware of these capacities, they adopt these capacities, they explore the potential to validate the value.

Next question comes from Andrew DeGasperi from BMP.

It's good to see everyone. And I just wanted to ask a question on -- in terms of your fastest-growing segments, Resources, utilities in the grid. And they could be construed as kind of related or indirectly related to data center spend and AI spend. And I'm just wondering, number one, would you agree with that? Number two, are you concerned that they could potentially slow in the future? And then, I guess, lastly, if it does slow, do you think that opens up capacity to work on other projects?

So 1 point of clarification that the resources is indeed our fastest-growing sector and then public work utilities in totality also grew strongly in Q2. So not just electric grid. It's not like the electric grid, we is growing necessarily much faster than other parts of the public work UTT sectors such as transportation or water utilities, et cetera. There is definitely a link with the big investments in data centers.

This is what's driving for self-sufficiency when it comes to critical minerals. This is also what's putting a lot of stress on the electric grid. So for sure, there is a link with investments in data centers.

With respect to durability, I think the need for critical minerals goes way beyond what is needed for data centers. And I think investments on the electric grid also needed way beyond what's also needed for measures. Yes. I think the of the need to secure critical minerals, they need to make the electric grid more resilient to potentially expand it, et cetera, that goes beyond data centers. So we are confident as well about the robustness of those end markets.

Next question comes from Alexi Gogolev from JPMorgan.

Thank you, Eric, and hello, everyone. Greg, I think in the past, you've talked about how tokens are trivial relative to engineering modeling and simulation compute and that much execution can remain local. So how does that shape your long-term gross margin profile for Agent workflows? And where do you see the main incremental cost centers?

Well, I think all the time, there's more awareness that things are going to wind up being hybrid choices. We would like those choices to be able to be made by our accounts and engineering firms and owner operators each have reasons to, as I mentioned, to ultimately favor relatively sovereign computing for their -- for reasons of their own sensitivities. That will be, I think, technically feasible as well with coming hardware advances.

And I think my answer is as far as its impact on our gross margins, it's relatively is likely to be relatively less impact on our gross margins as a result than perhaps others who are only investing in cloud-based computing requirements, because I don't think the world back to the previous question is going to be covered with data centers going forward and that there won't be reasons to balance out the computing form factors in a hybrid way providing providing AI is a good business, but that can include providing it in local environments as well.

The next question comes from Taylor McGinnis from UBS.

I'd actually love to hit on cash flow. -- it looks like in order to hit the high end of the cash flow guide, you have to assume pretty material second half cash flow margin expansion compared to it being down year-over-year in the first half. So could you just walk us through where you're tracking in terms of the guidance range, what would cause the big reversal in the second half of expense growth particularly given some of the AI investments and other areas?

Sure. Maybe I take that. So Tecteo, we are in H1 exactly where we messaged in Q4 '25 and our outlook where we would be expected to be so at 47% in H1 of our full year outlook, and we guided towards 45% to 50%, if I take the midpoint of the free cash flow outlook. There were 2 things we pointed out in Q4 last year that will lead to a shift between from Q2 into -- sorry, from H2 into H1, in 2026. One was that we had really strong collections at the end of 2025 which benefited 2025, but which was a higher comparison then for Q1 2026.

And then we talked about that we are doing investments into the business early in the year which also impacted the profitability and the cash flow in the first half of the year. So it's really just timing. It's reworking capital mostly. And it's quarter-to-quarter movements that really catch up quickly between what is the underlying model with our recurring revenues, negative working capital, low CapEx, all of that is still intact. So expect 50% to 55% of cash flow generated in the second half of the year, and we are on point to reach our free cash flow outlook.

I just jump in to say that on expenses and margin, you can count on us hitting the goal for the year because it's a fundamental incentive requirement for our executives, and we can manage to it and we do. anything that happens during the year is just quarterly planning, and it's not hard to achieve the goal because we manage to it. However, in terms of cash flow, we can't -- we've been -- I've been surprised to have invoices be paid before their due date in previous years, and that just take it when it comes.

And this question comes from Josh Tilton from Wolfe Research.

This is Arsen on for Josh. Just wanted to kind of ask if you could discuss maybe whether there's any benefit to including some more project-wise functionality, I think that was updated in E365 consumption and whether that's driving any higher consumption or supporting higher ceiling resets. And kind of with 3Q having that lowest level of sealing resets, -- you said kind of the same seasonal growth as fiscal '25 Q1 and Q2, but still marginally outperformed that both in Q1 and Q2.

And -- so when we're thinking continued momentum on an organic basis, should investors kind of expect a similar slight improvement in organic sequential growth rate on those sets.

All right. Let's -- a number of questions there as well. So hopefully, I won't miss any. First of all, the strong renewals was definitely a growth driver in Q2. It is our second largest quarter for renewals, but it's a distant second. The first 1 is Q4, but when it came to floor and ceiling uplifts, it was in line with what we've seen in previous quarters. So around 10%, which gives us a lot of confidence because this is always a reflection both of past consumption, but also what our accounts are seeing as the demand in their end markets.

So it's a clear vote of confidence for them to agree on these uplift. So that's what we've seen in Q2, and it gives us a lot of confidence for the remainder of the year. Okay. Now on project wise, I would say, benefit, the cloud overall is a growth driver. We introduced Connect at the end of 2025. And it is definitely a growth driver for us. The value proposition resonates a lot with infrastructure organizations. And this helps to support growth with both existing accounts and potentially new accounts as well.

Great. And the last question comes from Tom Zilberman from BofA guys.

Can you hear me?

Yes.

I think you can hear me, but you can see them by empty desk in the office has been working from home today. So I just hear my voice. I wanted to go back to the line of question on AI competition. but maybe frame it a little bit differently. I think since last quarter, of course, there's been the announcement around Prometheus. And I think also there has been a large European model announced a partnership with aerospace company for engineering simulation and design.

Nicholas, I appreciate your earlier comments that these are kind of different areas of focus versus your core competency in infrastructure engineering. But is there any risk that they eventually start converging to your area? And is that driving any sort of hesitancy from customers or kind of a desire to wait to see that these models improve and get better and eventually get to the area of infrastructure engineering.

Well, Prometheus is particularly interesting, and it's a bit skewed, which is fine. It's early stage. What I think we can say is such are the level of its ambitions as measured by its investment so far and what's talked about is that software and providing software tools can't turn out to be very much of what it has in mind given the relative size of that market compared to it's scale. However, these are conspicuous investments in the world.

And I think it does serve to help increase the enthusiasm on the part of engineers about what AI can do for them and what we can start to do now with their -- to provide them assistance. I believe that what these organizations develop will be useful and they include engineering functionality for their own purposes, but their own purposes are unlikely to be particularly competitive with our place in the market.

And to the other question with respect to our accounts hesitation because they hear things like some of these investments, absolutely not. This is not coming up at all. This is not the noise that I'm talking about. The noise is also their own explorations about what they can do with -- and what we're cutting -- and the way we're cutting through is just really demonstrating what's possible here and now without speculation, right? What we're offering is possible now and it's offering tremendous value for everyone.

That concludes our call today. Thanks for your interest and time. We look forward to updating you on our performance in coming quarters. Thank you.

Thank you.

Now, and Greg's already had one meeting. We're delighted to have Bentley Systems back here at the conference in London again, and have Executive Chair, Greg Bentley. Greg, thanks for joining us. It's wonderful to see you again, and so wonderful to have you here to talk about Bentley Systems. I think probably it would be great just to give a level set for everyone as to what Bentley Systems is and a brief overview of the company.

Thank you, Bob. So I'm the oldest of 5 Bentley Brothers, the other for all substantive engineers who founded Bentley Systems in 1984. I all have recently retired, including me, but we remain the majority economic owners of the company and the majority of the Board of Directors. So you should expect long-term orientation in how our company makes decisions.

We describe ourselves as the infrastructure engineering software company because we are the digital quarter masters, if you like, that provide the modeling and simulation tools and the enterprise collaboration environments for civil engineers, structural engineers, geotechnical engineers and those who work for them and by virtue of their work in terms of end markets, we are, we think, the leader in infrastructure engineering software for the public works and utilities sector.

And within that, for the subsectors of roads and bridges, rail and transit, grid and water, we're the leader in, we think, in the resources sector that includes mining and environmental modeling. And we participate also in the industrial infrastructure sector and the commercial facilities, infrastructure center sectors, but there we're not the leader, and you might describe that as vertical infrastructure.

We think we are the leader in horizontal networks of infrastructure, as I first described. Our business is -- reflects 42 years of penetration. There -- 2/3 of our business are with accounts that spend $250,000 a year or more in ARR for our portfolio of products. There's 200-plus among them who spend over $1 million a year with us.

Accordingly, we're 93% direct sales and our -- the majority of our mainstay enterprise ARR is through a consumption program where we charge per application per day, which also brings us transparency in our revenue and ARR accounting and in our margins and direct sales leverage helps us to grow our operating margins 100 basis points per year as we get more efficient to have reached about 30% in adjusted operating income less stock-based compensation, the way we measure that, and that flows directly into cash flows in a rather transparent way and we reached 35% in free cash flow margin.

In general, the basis of valuation, if you like, we grow those free cash flows in the mid-teens annually our ARR, we grow in the low double digits. And so the basis of valuation for our company has doubled every 5 years, certainly has 5 years since our IPO in 2020. And it's our job and expectation to continue to be able to do that.

Sounds like terrific financial performance. I think everyone in the room recognize that the market has recognized it and -- always not enough, Greg. We know that, right? So -- but it needs to catch up with what the reality is of how you're continuing to grow the business.

So -- and so can you talk through that financial performance and how you think about it and how you allocate between M&A, cash dividends, stock repurchases, how is -- how do you think about your financials?

Well, as a public company, we were able to do 2 what we describe as platform acquisitions of $1 billion magnitude in 2021 and 2022. We took on some convertible debt to do that. And that -- if you count the convertible debt as debt, our leverage was at 5x or so. We've worked it down over the past few years to 2x or so and are at equilibrium now where our priority for capital allocation would continue to be programmatic acquisitions.

We'd like -- we're always able to do another platform acquisition, should that come along, but it's very fortuitous when that might occur, and we can't plan on that. We do make sure to repurchase stock at least to make up for what otherwise would be dilution from stock-based compensation which is important to a company like ours, but is reasonable in our case. And in recent years, we have repurchased about than last year about twice that much stock, given the level of our stock price. And we also pay a modest dividend.

Wonderful. So we're not going to get into politics, but there was a bill that recently passed out the Transportation Committee called BUILD 250. And so that's moving through the house. what could this mean for Bentley?

Well, during the -- the Infrastructure Investment Jobs Act in the U.S., which was the first serious long-term infrastructure investment at the federal level in the U.S. But that's almost 5 years ago now. And so that bill will expire during this year. There's still a lot of money to be spent in the future from it.

But during that period of time, we and the engineering community urged that the approach to think about in the U.S. would be longer rather than faster for infrastructure spending. So it is reassuring that this bill, which so far is only in one part of the legislature continues the level of funding with slight increases for surface transportation. That is the job of this committee that's released this bill.

So surface transportation is highways and bridges, but that's been traditionally what's funded at the federal level in the U.S. So it's good to have another 5 years rather likely to -- there might need to be a short-term extension of existing funding, but that looks like the future for surface transportation, which is a good thing for us.

The other areas covered under the IIJA that will soon expire where federal funding for broadband grid and water in addition to surface transportation. But those areas -- in those areas, there was no lack of private funding in the case of broadband and grid and water is funded more so at the local level in the U.S. So those sources will continue.

That's not the remit of this new bill, but no one's concerned in those areas, especially a broadband and grid. Grid especially the concern is not funding levels, but permitting levels. And this bill does what it can to address permitting obstacles for surface transportation, but there still needs to be permitting reform in many countries, including the U.S. to be able to expedite investments, especially in the electric grid and especially the transmission high-voltage portion of it, which will be much to our benefit when it occurs.

Yes. As Americans, I think we can both agree that there's a lot of infrastructure that needs to be built and Bentley will be right at the center of that as it continues to grow.

So it's taken me the question 4 to get to the thing that everyone likes to talk about, is that AI. So can you help us understand how Bentley is positioned in this age of AI that we talk about every day?

Well, the positioning for AI, the -- what first comes to mind is how zealous and enthusiastic we are about the new opportunities this opens up. It literally is the springboard for our new generation of management. And first to say in terms of positioning our -- we don't think we're particularly vulnerable in what we do in software to incursion from AI.

We don't have any administrative software that might theoretically be vulnerable to being self coded. And engineers modeling and simulation software is the way in which they conduct their work and is subject to their compliance responsibilities and personal and organizational, professional and legal liabilities, that work is not going to be done by software that's probably right. It has to be provably right.

But in terms of opportunity, which is the way to think about it and sort of in order of revenue immediacy, AI opened for us, a new opportunity we call asset analytics which is instant on AI for owner operators of infrastructure. The owner operators of infrastructure are half of our business, the other half being the engineering firms who are their supply chain.

But for owner-operators asset analytics is a new process of using drone flights, especially and other sensing and surveying methodologies to automate the inspection and monitoring of their infrastructure assets. And then the next stage -- the next step is to apply our engineering modeling and simulation applications to derive engineering insights to improve the operations and maintenance.

And so that's a business that's underway and so far has reached recurring revenues of about $50 million for us. And there's a litany that I'll shorten of AI initiatives in our company as with any company, but that would include new products that are AI native and are percolating through but capabilities from those products that we're extending throughout all of our applications, for instance, to automate drawing production from engineering models.

But the perhaps the most exciting one that we can talk more about is the opportunity to provide agentic expansion of what an engineer can do to improve the quality as well as the quantity of their work with agents that are on their behalf rather than as a substitute for the engineer.

Good. So let's kind of stay on that same theme because it would seem that what you're talking about is engineers being more productive using AI. And so how do you think -- how do you see that impacting your seat-based consumption and kind of your overall ARR?

Our application consumption should increase substantially with AI and -- the thing about infrastructure engineers is there is severe resource constraints on their capacity, and they can't do any more work. It's very important to make them more productive. And for instance, just saving them the tedium of drawing production will enable them to do 20% more projects that tackle the backlog and allow their firms to bid on more work finally.

But it is -- they're supplementing their attended consumption. What we currently charge for today, they sit in front of a computer and instruct it. When instead they spin up AI agents on their behalf, we get to terrific new capabilities to improve their work. And the model under which they work, the commercial model generally is that they're being charged by the hour, infrastructure under operators are paying for infrastructure design for the hourly time of engineers.

That means that a project minimizes the amount of time spent and has a limited budget during which the engineer can produce one iteration of the proposed design. With AI, their agents will be able to run in the background and optimize multiple iterations, hundreds or thousands of iterations in our first MCP server for structural engineering software in the first production project, we tried, the result of optimizing was to reduce the material cost by 40% and through this optimization process enabled by agentic AI.

But across our portfolio, I envision that the -- in addition to optimizing projects for cost and materials and carbon, we'll be -- the engineer will be able to have an agent which will optimize the constructability by running in the background while the design is being done, our SYNCHRO 4D modeling software, the sort used to construct these buildings where you simulate the occupancy of space and time and be able to deliver a project that's constructible as well as optimized in terms of cost to be able to have an agent that looks in project-wise, the system of record for prior designs and which is AI ready for such searches by virtue of our iTwin schema defined modules that can be reused that would improve the performance for an owner because they'll be maintaining consistent kit.

You would have an agent that could be investigating the subsurface conditions with our sequence software because otherwise, the risk in ground conditions is what causes project overruns, Ultimately, through Bentley Infrastructure cloud, the AI agents will be able to look at the performance of comparable assets and recommend those components, which have been most maintainable, for instance. So overall, there will be a much better quality of work that's possible, that will help the engineers charge appropriately for the value of their work.

But you don't see this as a risk that your customers will use the AI agents themselves or build them themselves and reduce their overall spending with you.

The -- our accounts should absolutely be encouraged to develop these agents themselves, the more that they do, the better. They have, especially the largest accounts with whom we're working most closely our E365 teams that are deployed within these accounts are helping them to introduce these agentic AI workflows now the -- we don't charge for them at the present time. We're only charging for attended consumption.

The much greater API agentic consumption, we will charge for in the future, but our approach is to proliferate it as much as we can to have the value understood and appreciate it and assessed and then we'll work out in an acceptable way to a mutually acceptable way to charge for that in the future. But the -- our accounts should allocate their AI development to such agents, which start at the top for what it is that they optimize and how they do the trade-offs and brand and market that to owner operators.

They have little incentive or opportunity to start at the bottom with the fundamental engineering logic that is already tried and true and for which it would take years to have some substitute replace that. The -- if we talk about the economics here, what we charge for our software in E365 is about 3% of the cost of an engineer's day.

What -- in a project overall, a design project about 1% considering everything else that gets done besides engineers software time is spent on software. The total design is about 5% of the total installed cost of an infrastructure project of the trillions of dollars spent in every country for infrastructure projects annually, or if you work it out, 1% times 5%, 5 basis points get spent on software and computing.

With everything I've described with agentic AI, incorporated in the project on behalf of the engineer it will be well worth spending multiples of that to get better performing infrastructure with less project risk. And so the opportunity with AI for us is orders of magnitude, more spending on software and computing more comparable finally to the rest of the economy that isn't slowed down by this time billing commercial model.

And with trillions of dollars being spent all over the world on huge infrastructure projects, it seemed that you're positioned well to capture so much of that.

It's a really worthwhile thing to work on. And AI is bringing forward the future for us, if you like. The things we have wanted to do with our software are more feasible and in front of us at the moment, and it's really exciting to do that.

Over the 42 years of our company, we've reinvented the technology underlying our software meant multiple times, that continuity in doing that is what has made us successful because of the longevity of infrastructure design projects and infrastructure assets, which is forever, to be able to have that continuity and yet incorporate new technologies such as AI. AI is software, by the way.

Yes. So let's talk about the concept of headless software. It's gaining attention, obviously, due to AI. Is Bentley at risk of losing mind share with customers if your software is increasingly being accessed through an external LLM provider?

Well, the LLM are going to be great to devise and run these agents to direct this process, but the value of the process is in the underlying modeling and simulation software. And no, we're not worried that highly educated engineers are going to confuse what's doing their work and livelihood as they already know and trust Bentley. So this is not a concern for us. Headless software is great and better. The more there can be, the happier we are.

So one of the things that we've heard a lot about, I think, just probably in the past 3 to 6 months has been software companies have announced layoffs to fund AI and some of you already spent their annual token budgets. I think that's probably been the talk of the past couple of months. How is Bentley thinking about funding AI investments and how may that change your overall cost structure in the future?

Well, our CEO, Nicholas Cummins says the proportion of what we spend in the future on R&D will go up. We'll have more developers, everything else in our company will get more efficient, faster with AI, and there's no end to our appetite to go on the software development road map that we can tackle much quickly at a much higher, much faster pace now.

The cost of tokens to do that is not going to sneak up on us because we already manage our R&D, what we call head cost. So head cost is not just compensation. It includes the cost of cloud consumption. So we're already considering and bearing this cost while we continue to grow our development resources.

I might point out that over our 42-year history, every engineer has the best, most capable workstation that you can imagine. And traditionally, these have been graphics workstations and the G in GPU computing means graphics. Software developers will have lots of local resources to run their trained coding models.

And I think our engineers, especially will because they will be testing the software. AI is software that our users will have a choice to run in a hybrid environment these optimizations that I'm describing can often be run locally on sovereign AI environments or in cloud computing for -- when that performance is needed.

But we should remember that the infrastructure design data is sensitive and confidential and needs to be secure. And the same is true of the intellectual property of engineering firms and owner operators, it's -- their data is their property to be -- to train their models for their purposes. It's not ours or anyone else's. So there will be actually a premium on hybrid environments and sovereign computing.

Yes. I'm sure that's clearly the case. And as people worry about what's happening in the world of LLMs to their data and how it's being monetized by others in the future. So resources have been a nice growth driver for the business as of late, as you mentioned. What are the drivers? What are those drivers? Are they susceptible? And if they -- excuse me, are they sustainable? And if the sector continues to grow faster than the company average, does that induce more cyclicality into your business?

Well, resources is our term for the sector that includes mining upstream oil and gas, environmental modeling generally. And this year, that has been a contributor to our strong start for the year, in ARR growth terms. We might have some concern about cyclicality historically in mining, but our sequence software is in the main used during the operation of existing mines where every scoop of ore brings you new information to improve your 3D modeling of the subsurface.

But we do have -- we do benefit from an increase in new mining activity. The thing is this may not be a typical upswing, but rather has a secular motivation, a secular driver, which is the determination by each country in the world now to become more self-sufficient in terms of their supply chain and especially critical resources.

And we can suppose that will continue to be the case foreseeably given the geopolitical situation in the world. A further driver for Seequent, our subsurface modeling company is the opportunity for all civil projects in the world to be ground informed because that's where the risk traditionally has laid. We've quadrupled that aspect of Seequent business over our 5 years of ownership, but it's still the case that only about 30% of projects include the subsurface modeling.

New data centers are starting to change that in the way they think about their potential need for water resources in the future. And anyway, there's -- this is a very strong sector for us. Our Seequent overall is growing on the order of twice as fast as the company. And generally, applying 3D below the ground. And why is it different than above the ground? Because above the ground, we can see and sense and survey, but below the ground, we can only sample with drill holes and boreholes, and then we need to interpolate and understand the -- with mathematical geological models, what lies below, but that -- applying 3D modeling for that is still a frontier that is expanding to our benefit.

So these are reasons I think it's generally okay for our company because we have such a flywheel, as I've described, our role, the number of engineers isn't changing much because as many as are retiring as our joining the workforce and the new ones are generally in India and China, but they're all working full out.

And that's rather predictable in terms of our traditional business, ARR, and it's appropriate and okay for us to take on some volatility. Asset analytics is of that nature as well where we compete for large procurements in the inspection and monitoring opportunities for major utilities across, for instance, their fleets of cell towers or distribution poles, for instance, where we've started. But we -- it's a good idea, I think, for us to take on some volatility at the margin in order to place our bets to grow faster.

So you touched on a topic that I think is of interest to all of us these days, which is data centers. So can you speak to how Bentley is positioned for the build-out and power needs of data centers these days?

Well, first of all, the data center market is one reason that engineering firms are as busy as they are. And so confident about their backlog and future work, which is reflected in their commitments to our E365 program with floors and ceilings to bound their consumption in future years. So we benefit overall from data centers, but from -- in a broad way, but specifically for data centers, while they're vertical infrastructure, if you like.

So the design itself is likely to be done with other software. Our structural engineering software is probably used in each case and our geotechnical software for the foundations in each case. But the need to construct them as fast and efficiently as possible is requiring our SYNCHRO 4D modeling software, the same one used in these buildings to build in London with no layup yard possible. That technology is being used in data centers as well. So that's our first level of benefit directly.

Next, each data center is a campus requiring utilities. So the roadways, the water networks and so forth are where our software is used in each such case. But then even more broadly than that, at the macro level, data centers increase this need for critical minerals and copper and ultimately, water, including aquifers and geothermal energy, where our sequenced software is always used. So it's a contributor on all those fronts.

Yes. It sounds like there's a nice flywheel effect just in that segment of the business for you. So has been a strong growth driver since the time of the IPO, so 5-plus years ago. How much more runway do you think you have for that to continue to be a strong growth driver for the business?

Well, we didn't explicitly address small and medium businesses prior to our IPO and our new management appropriately have approach this again through direct sales and increasingly digital. We never did an e-commerce transaction before going public 5 years ago, but it's become a strong contributor adds 3 basis points of ARR growth reliably and over 600 new logos for each of the past 18 quarters now.

What we are on our learning curve focusing on now is the upsell and cross-sell opportunity to SMB accounts who have joined us -- it's still perhaps 1/4 of our business and the -- their turn out to be as many infrastructure engineers who work for small, medium businesses for us, we classify those with 50 engineers or fewer. There are as many such engineers and no smaller firms as they are in the larger firms.

So you can see that there is continued opportunity for us there and reaching them digitally, they prefer to do self-service. They don't particularly want to talk to a salesperson. They're ending up acquiring software. And by the way, it can be even more specialized software, the more -- the smaller firms can be more specialized firms and use more specialized and expensive software subscriptions, but they don't particularly want to talk to a salesperson, but our inside sales people are engineers themselves and we provide engineers as well to help them use the software.

So when they first say, "Gosh, we know of the reputation of Bentley Systems software. We know the large firms do the large projects with this software hadn't thought that I might drive with Bentley dental, but actually, if you'll help me, it's worked out pretty well for them, and we do have considerable remaining upside there."

And so there's still plenty of small and medium-sized businesses all around the globe to tap into.

That's where half of engineers work, and it's very worthwhile for us to have now focused on that opportunity -- that incremental opportunity.

Great. When you think about the other growth drivers, you've talked a lot about getting bigger in the operations and maintenance phase of infrastructure. Why is this an attractive market for you? And what are the capabilities that you have today and where are you going with them?

The opportunity for digital twins in infrastructure is an opportunity largely to improve and optimize operations and maintenance. That's the -- that's where the bulk of spending occurs. That's where all of the benefit occurs during operations and maintenance. And the norm at the present, unfortunately, is that the work of the engineers is only utilized during project delivery and never thereafter.

So to have an as operated engineering model can improve the resilience and safety and life cycle performance of an infrastructure asset, but it used to be hard with AI and our asset analytics approach, we can stand up a digital twin overnight with drone flights and other capture of that sort we started with cell towers because they're owned by organizations that are private.

They're providing the broadband for the rest of us, and they're inclined to be progressive toward technologies. From the drone flight you have video overlapping imagery, our software processes overlapping imagery into a reality mesh model that is engineering ready and you actually perform engineering, but you then bring in the engineering models because the first thing you do is identify whether the cell tower can accommodate more equipment and therefore, generate more revenue.

But each such piece of equipment would require that the wind resistance be remodeled and the structural sufficiency be calculated again an electromagnetic interference and so forth. So to be able to have the AI and computer vision be supplemented with the structural and other engineering models is the approach that is launching us to this large opportunity to address, and we now do distribution poles and we'll go on to transmission towers and other such structures, roadway miles.

These will be large business opportunities that will be gained by someone. It would be us, we think, because we incorporate the engineering and logic and the engineering models to get insights from the survey modalities. This is our preference for approaching the operations and maintenance opportunity. The driver is the opportunity for owner-operators to spend less on maintenance because they won't do -- they won't need to do guide book maintenance and often they have to defer that because they can't afford it.

If you continuously survey, you -- it's more economical and you avoid dangerous things for humans to do. You can observe what maintenance is working and when maintenance is required and because you're continuously monitoring and rerunning the structural sufficiency and so forth, you can do only that maintenance, which is working and is necessary and spend less and the digital twin can cost less than nothing with asset analytics.

So that is truly a tremendous opportunity, we charge per asset. So it's a whole incremental business model in addition to our traditional business of charging per engineer. And we're -- this is our preference for how to tackle the opportunity in operations and maintenance. We wouldn't, for instance, acquire administrative software for work order management and so forth and immediately become antagonistic to the incumbents who provide that already for infrastructure or owner operators. We'd rather integrate and add engineering models through digital twins in this respect.

So we began our conversation talking about the Bentley family and you and your brothers. And a couple of years ago, you moved to Executive Chair and brought in Nicholas as your new CEO. What are you most pleased with that you've seen with him over the past couple of years? And what should we continue to expect from them in the future?

I'm so pleased. It's hard to call out something, but I'll say balance. We -- in our generational succession, the first time in our company I've been CEO for 30-plus years. We wanted a balance between left and right brain. We've always done better on engineering and less well on marketing, for instance. And we focused on large accounts and left out SMB in the past and so forth. And I think we've really succeeded with in that.

Another aspect of balance is in executive talent recruitment, Nicholas has really recognized that the importance of what we do in the world. And now that we're a public company with a bit higher profile, we can and, therefore, should attract talent from around the world. By the way, Nicholas is based here in France -- in Europe, in France. And he has been not settled for less than top global talent that continues to please me and the opportunities we talk about with AI are in great hands.

Great. Well, we do have a few more minutes in case there are any questions that we want to take from the group.

Thank you very much for your attendance today. This live stream has now concluded. Thank you.

Okay. Great. Hi, everyone, and welcome to JPMorgan Boston TMC Conference. My name is Alexei Gogolev, Head of Vertical SaaS team. And today, I'm delighted to be hosting one of the key founders and majority holder of Bentley Systems, Company Chairman, Greg Bentley. Greg, welcome.

I guess first, Greg, I wanted to start by discussing the fact that the stock has traded as if the market is somewhat skeptical about Bentley's ability to fully capture upside from AI? And in many of your conversations, you've noted that the market has sort of discontinued and discounted your value in the near term. So what's your best case for why Bentley is positioned to be an AI leader and perhaps a long-term winner in this space?

Well, Alexei, the very long term is how we set our compass Bentley Systems, and it perhaps is not surprising given our family majority ownership. And that's the case with our generational succession in management. I was CEO for 30-plus years. Our new management team, new to their positions but not new to the company necessarily are in their 40s. And their refrain is infrastructure AI as their opportunity. And I had thought of that as rather audacious in the scheme of things, such a major opportunity is that. But as things developed, I've come to be astonished at the pace at which AI is bringing the future forward for us and what we have always wished to and aspired to bring about better and more resilient infrastructure in the world. And that includes the operations and maintenance of infrastructure and opportunity, which is relatively newer for us.

And of course, by -- in bringing the future forward, it's the case that AI accelerates the pace of our own development on our road map to the future by multiplying the productivity of our software development, but I'm mainly talking about the inception of changes in the market structure for infrastructure that are truly auspicious for us. And I've been talking this quarter about engineering firms, top design firms in the world because they are together our largest constituency. And I was delighted this week to see the CEO of our largest account in his own comparable transcript say that, well, on the one hand, AI and technology is important to them because their backlogs are at record levels as with everyone among our accounts. They can't hire more people collectively in the industry because the number of infrastructure engineers only increases 1% per year.

And that can't change foreseeably given the demographics of those retiring in comparison to the number in school. And that therefore, to do more work, they need to be more efficient and AI can help with that. But in response to an obvious question about whether AI will reduce their firm's work and revenues because their business model throughout infrastructure engineering is charging per hour for their work with the efficiencies of AI, will they have lower revenue? And he said, I don't believe a bit of that. Our design work has not been optimized yet.

And this is the opportunity we have been identifying with AI for engineering firms that in addition to the attended consumption where an engineer sits in front of our modeling and simulation applications, they will now, with agentic AI, be able to spin up agents that run hundreds and thousands of iterations of their potential designs to optimize in various dimensions and generate better and more valuable products for owner operators, and he was talking about that value, the risk mitigation and so forth. In our own portfolio at Bentley Systems, these levels of optimization include across multiple disciplines. It includes optimizing the constructability of a project with our SYNCHRO applications.

It includes mitigating the risk of subsurface conditions, which is the principal risk that occurs in an infrastructure project, optimizing to reuse prior design modules to derisk and improve the quality and then ultimately optimize the performance of the design by being informed by the performance of other designs for the same owner and how well they have performed in the field using our Bentley Infrastructure Cloud data about operations and maintenance. So all of these opportunities are coming together, thanks to AI and when the design portion of an infrastructure project has broadly averaged about 5% of the total installed cost and the portion of design billings spent on software has been about 1% of the design billings. That's 5 basis points of the infrastructure project cost, the infrastructure projects, which collectively across the world consume trillions of dollars, new dollars every year, 5 basis points of that is currently spent on software to make the design more intelligent and optimized and so forth.

The CEO's point was when we're being paid by the hour, we only have time to do one version. If agents can help us, we'll be able to do a much better design. We'll be able to be paid for that. The owner will not be interested in optimizing and reducing the 5 basis points spent on software in the project when it can be a derisked project and perform much better. So now over on the owner-operator side, the opportunity for optimization in operations and maintenance is even larger. The -- what we call the digital twin opportunity is that the engineering logic that the owner pays for during design is at present, never used thereafter during the remaining operating life cycle of the asset, where 80% of the cost occurs in that maintenance.

So to be able to optimize maintenance would be able to do only the needed maintenance and our asset analytics opportunity uses drone capture of existing conditions typically, among other capture modalities to have an as-operated calibrated model to which you then apply the engineering logic created during project design to make decisions about whether maintenance is needed and to be sure that the infrastructure remains safe. So those opportunities accelerated by AI asset analytics is only made possible by AI have come together to create an economic opportunity for better and more resilient infrastructure in the world and get past the inhibitions and inertia that the existing institutional business model of charging by the hour have caused in the past. So at the same time as we have the technical capabilities from AI, we are getting past the commercial model obstacles to smarter and more resilient infrastructure.

So much more will be spent on software and computing and infrastructure engineering and I can't make the claim that, that will all be to the benefit of Bentley Systems. But our position is kind of ideal for this. In our 42 years, we are the trusted quarter master, digital quarter master for the engineering organizations and the owner operators, each are half of our business. Our modeling and simulation applications, the ones you need to invoke agentically in order to optimize are the established standards in the world and trusted by the engineers. And last of all, the data needed to do this optimization is in our project-wise environment. It is the intellectual property of the engineering firm or the owner operator or both, but it positions us very well to help them themselves take advantage of that proprietary data in their proprietary differentiation to improve and make infrastructure better and more resilient.

Greg, you released the STAAD MCP server and said you're exploring token-based pricing for API consumption. So can you give us some early data, how many accounts are actively using the STAAD MCP server today what does a typical session look like in terms of compute intensity and when should investors expect the first commercially priced API consumption contract outside of asset analytics.

Well, it's only a couple of weeks now since this first of our MCP servers was released. And over the coming year, we'll completely outfit and instrument our application portfolio, both modeling and simulation applications. So there's not a record yet, there are more than a couple of hundred downloads, and we don't think we can count those that are done through Claude itself. But what we can answer is the monetization question, which is that we'll have no monetization this year from API consumption or MCP consumption because our determination is to benefit the long term by having our accounts do as much exploration and experimentation with the APIs and MCP services as they can do with no intention to monetize it during this year. We reckon that this year, we'll learn for instance, about the cost, as to which you asked.

And the reason there's not a big risk to us in this year is that initially, the APIs and MCP services execute on the engineers on desktop. And that's a machine they have available to them anyway. It's a terrific -- they find that they can do hundreds of iterations at machine speed. At the same time, they could have done one at human speed. And we'll hopefully get experience as well and what the benefits are of these optimization approaches as we enable them to become richer and richer with more and more APIs across our modeling and simulation applications during the remainder of this year, Alexei.

Okay. Thank you, Greg. So at the most recent earnings call, you've laid out really well the math on representative, let's say, $1 million design project and how software today is about $10,000 in that contract, and you suggested that hypothetically doubling it to $20,000 with API consumption could really save almost like 20% of engineering labor. So that looks like a very compelling ROI. But what's the realistic adoption curve? Like are any of your hundreds of accounts spending over $1 million already and already running some agentic workflows in their production? Or is this still entirely in pilot?

Well, we have 220 accounts that spend over $1 million with us, 820 accounts that spend $250,000 and more. And these are big enough enterprises to have been already investing in AI R&D. What they don't need to do with our agentic API road map is start at the bottom replicating functionality that they couldn't replicate and wouldn't be able to validate any way for modeling and simulation, but they can put that together into levels of optimization all the way up to the project level with the context they know about, that's a wonderful thing for them to be doing. Many of them are already doing that, exercising our APIs that exist in our products today without our help, we don't have a way of tracking that. We don't have a way of monetizing it, but we want to encourage it anyway for the reasons that I just mentioned.

Okay. So in terms of certain bear case scenarios, it sounds like it could be that horizontal AI platforms with agents having access to open source structural analysis tools could eventually replicate some of that STAAD and what it does without paying Bentley for API access. So would validated simulation engine and the regulatory requirement for proven tools or maybe something else prevent such a displacement?

Well, you're asking about our first MCP service is for STAAD. STAAD is our structural application that is, I think, the leader in the world in structural analysis, and there are open source structural analysis programs now that our accounts could be using. The STAAD users are not, in general, they prefer the commercial products, which cost them only about 3% of an engineers cost each day on average under our E365 program across our application set. So there's not a great incentive to not need the best product set. STAAD understands the design codes, which the open source ones don't in their jurisdictions, it does some optimizing at the element level. So its functionality is better. But generally, the engineering firms have much more to gain by spending their AI discretionary R&D investment on leveraging the existing tools that are proven, that are mandated by their owner clients and which the engineers already know how to understand and use and have chosen in these levels of optimization.

And what I mean by levels of optimization, you could optimize material or cost for a structural solution, but a structural solution is only one of the disciplines that need to be solved. And a design if you could orchestrate across those to have an objective function across the design, that would be good. But for a project, you might want to also optimize the availability of the materials, the type of labor and maybe conformity with designs and assets the owner already uses, for instance, these are all aspects of adding value that we can't do in the granular applications? And is the opportunity preferably for both us and the account for them to add in optimizing at a higher level of value.

How about the fact that you've highlighted that you explicitly need consent from your customers to train your own AI and those models. Obviously, some of your competitors may consider using their customers' data without consent to train their models could your principle stance become a competitive handicap rather than an advantage longer term?

Well, our principal stance is that our users intellectual property is their intellectual property, and we're not entitled to appropriate it to co-mingle it to share it, especially not to have it help their competitors. And that may seem different than other software companies viewpoints on the matter, but our position is a bit different. For one thing, before we get to the principal of it, consider the sensitivity of this data. This is -- these are the designs of the world's critical infrastructure and consider whether anyone would want uploaded to cloud or downloaded from ChatGPT, whatever, the vulnerabilities potentially for terrorists or adversary, et cetera, of the airports and the wastewater and grid and so forth can't be anonymized by the way, because it's intrinsically geo-located, each such design.

So we have a responsibility to secure it and steward it. But that doesn't disadvantage our account whose data this is because it's sufficient data for a given owner of operator of infrastructure, they're all large. There's lots of data that they have. In the case of engineering firms, the same is the case our opportunity is to help them make the use of -- make the most use of that data. An example would be, for instance, in automating drawing production, which is where 20% gets saved of the engineer's time that allows them to work on an incremental project at a fixed pricing and improve the margins very substantially. But automating drawing production can use the standards of a particular engineering firm or perhaps a Department of Transportation would want all of its drawings to a particular dialect that works for them so that they're the contractors to whom they're sending those drawings for construction bidding and can expect a certain consistency.

That training can be done sufficiently with data, which is the intellectual property of our account or which has agreed to share it. So we do not even solicit permission from our accounts to have standing permission to access and use their data for AI training. When we have particular projects and to train models, we can ask and be authorized to access certain data that is not differentiated or confidential in the accounts judgment. And then we have a data agreement registry to keep track of that. But it will serve us very well because of the scale of these accounts and the amount of data they have to help them in a proprietary way, improve their differentiated offerings by using their own data where we help them do that. That's the right compass setting for us, we think.

That makes all the sense, Greg. Talking about the -- one of the attractive businesses within your portfolio, asset analytics, it recently crossed $50 million revenue run rate. You've been clear that not all of it qualifies as ARR because inspections aren't always annual. But can you maybe help us size this more precisely longer term? Like what needs to happen to make most of this revenue annual is it some regulatory change that you're expecting or maybe some pricing, packaging or some technology change?

Well, to back up a bit, we have forever been promoting the potential of digital twins, which reuse the engineering intelligence developed for the project delivery in the first place after delivery during operations and maintenance. And in parts of the world like Asia, that's taken up right from the outset. Our new metro system will have a digital twin approach right from the start and use -- and if it's done in Singapore, that will be the case, for instance. But and owner operators, for instance, in the U.S., that's too complicated. They don't have many new projects starting and what asset analytics enables is using new continuous surveying, data capture technologies, drones in the main. I always do this wagging my finger to show a drone, circling a drone produces video.

Video is overlapping images. Our software resolves overlapping images into a 3D reality mesh that captures the digital twin of the operating asset today and then you make it intelligent with AI, recognizing everything and relating it to the engineering models. So that's that enables asset analytics to create an instant on digital twin which you can use in operations and maintenance to know what operations and maintenance is required and to ultimately optimize that. So it's a very powerful opportunity. We started with cell towers and then roadway miles from Dash Cam data and Google Street View and now distribution poles, for instance, that together add up to this proven $50 million.

Now it is recurring revenue, but the -- and it's a subscription, an annual subscription in each case, but it may be a subscription for a year when construction is occurring for a particular pole or a tower or asset or mile. And it may not necessarily be renewed for subsequent years. In the case of distribution poles, an inspection is required every 5 years in the United States. But so our hesitation in classifying it as ARR, it is recurring revenue, it may not be annually recurring if you don't renew it every year. it may be somebody different that you need to sell to after the construction or after the inspection. And we're working on that.

The best way to work on that to make sure that the surveying by drone or otherwise is, in fact, continuous and occurs regularly is to make it simpler, have more AI extract more value from it and make it more worth doing more often, which in turn, by the way, if you knew you were going to inspect something regularly, you wouldn't need to replace it quite so often because you knew -- would know you're going to be able to catch a problem before it becomes significant, which the engineering model would tell you. At any rate, you can imagine also in terms of the ARR challenge for an account having floors and ceilings because they may be doing different assets each year and so forth. So we're doing all of those things to make it an ARR business substantively and contractually. But the most important thing is to capture as many assets get out there in front of this big opportunity that we can do.

Greg, since acquiring a business called Seequent, you've grown subsurface ARR quite materially in civil infrastructure. I think you were saying by a factor portfolio 4. I don't know what dollar amount that it equates to, but what's the addressable opportunity for that division? How many of your top 470 design firms accounts are already using Seequent for ground investigation?

Well, Seequent was what we call a platform acquisition almost 5 years ago now. And it is a global company that started with 3D modeling below the ground from mining, in particular, we saw the opportunity to apply that technology which is different technology below the ground. You can't see and measure and take pictures. You need to do boreholes and samples and drills and then informatically what the strata are below the ground and where the water is seeping and environmentally modeling of that, we saw the applicability of that to civil infrastructure projects where ground conditions are the biggest source of risk that cause project overruns and the Mining and Resources business has also turned out to be good.

The Seequent business has multiplied while we've owned it. But the civil portion of it has grown to be a very significant double-digit proportion, but there's lots of headroom there as well. It happens that not quite half of our top design firm accounts that you're referencing use Seequent at this point for subsurface modeling for ground informed design for derisking projects and they all should for all their projects, and we'll work -- continue to work on that opportunity.

Perfect. Switching to E365. I understand it's now 46% of ARR and growing pretty rapidly. Can you walk us through the unit economics of E365 renewals? And what's the typical floor uplift you're achieving at those renewals? And how does that compare to maybe a few years back?

Well, E365 is a pure consumption program, we're paid for and our revenue and our ARR are strictly based on for our open applications, the number of days used for our project-wise environment, which is 20% or 30% of it, it's per quarter in that case. But what we're -- the great aspect of this is there's no multiyear contracting and commitment or accounting. It's strictly ratable revenue recognition back to the good old days before 606. And what occurs at renewal is an annual price escalation becomes effective. So there tends to be some increase in ARR. But we were asked by accounts, although this is not an aspect of the program contractually, if we would grant them a ceiling on what their consumption could become in the ensuing year.

And we said, yes, if you'll grant us a floor. So a collar became conventional for the contract to be negotiated a contract renewal. And then a few years ago, accounts who saw this sort of thing elsewhere said, can we negotiate that for multiple years in the future so that we'll have a cap on what we could potentially spend based on how our consumption goes over time. And we said, yes, if we could symmetrically negotiate a floor as well. And that has become the norm for each of these renewal negotiations. What happens at renewal mechanically is our ARR would be increased by the price escalation, which tends to be in mid-single digits each year. The remainder of the consumption increase is there. And the main there upsell to use more specialized applications of ours, a better for us, mix of product usage.

Not much of it anymore comes from increased volume because there is a limited number of infrastructure engineers as we said earlier. But there's information in the result of this negotiation and I always ask each quarter, where did we wind up with the average of the negotiated collar increased escalation out into the future and the answer for as long as we've been doing it is 10% or so on average across the world and across all account sizes and so forth, which is the mentality. I think we can gauge from that infrastructure engineering organizations, the substantial ones. These are the ones the 2/3 of that pay us $250,000 a year and more are confident in their need for an expectation for and what they'll actually pay us is what they actually consume.

It's a very fair -- the most fair contract format you can imagine, but they're confident enough about their next several years, notwithstanding AI concerns and threats and so forth to be able to commit to floors and we grant ceilings at that 10% level. So that provides a lot of consistency in our ARR, it must be said, I take the view as an owner of the business that it's so much consistency allows us to take on some risk on the margin like the asset analytics business, which is less predictable going after big contracts and so forth. We're going to have a happy medium while we work on building that one up ultimately to be comparable, I think. But in the meantime, AI is on everyone's mind, and it's not a deterrent to continuing to commit to the level of attended consumption because that's what we're charging for now for our existing applications.

Perfect. Greg. And final question. So you've talked a lot about having capacity for up to $400 million worth of annual programmatic acquisitions without increasing leverage. Last year was a little bit light, so you spent less than $100 million on acquisitions. What do you feel is holding you back from deploying more capital aggressively? Is it lack of targets, maybe some valuation discipline or integration bandwidth and how that -- now that you've expanded beyond asset analytics, what other domains are you looking at?

Well, for a period of time after our platform acquisitions, that's the billion-dollar scale acquisitions of Seequent, which was, at the time, about 10% of our size and then Power Line Systems the next year, we issued convertible bonds that if you consider them, that brought our leverage ratio up to 5x or so and over several years, we've worked that down now to be under 2x on which we regard to be tolerably optimal. So we have no wish to delever further. And that gives us the number given our cash flow is above $500 million this year -- last year and this coming year to be able to do $400 million or so of acquisitions. That is our priority. We would like to especially acquire AI for additional asset types and asset analytics beyond the ones I mentioned, just to get the time advantage of not needing to take the time to do the training ourselves but they're scarce to find those good assets.

And it took us until the end of last year to do the couple of acquisitions we did do all that was our priority. But now while we're -- we've reached an optimal leverage. We will continue also to repurchase our shares. But the greater priority will be such acquisitions, including now since we can expand beyond the asset analytics opportunity, where we are disciplined, and it's not the case that we're allergic to dilution in margins compared to our very good margins but we are disciplined in valuation compared to private equity firms perhaps still. And we are now returning to prioritize as well acquisitions of established simulation engines, especially that have the value in this agentic immediate future that I'm describing of allowing us to optimize on even more dimensions of engineering performance. So you'll see, hopefully, both of those, we really do hope so in this year.

Perfect. Greg, thank you very much for joining us today. I appreciate your time.

Sure, thank you.

Good morning, and thank you for joining Bentley Systems Q1 2026 results. I'm Eric Boyer, Bentley's Investor Relations Officer. On the webcast today, we have Bentley Systems' Executive Chair, Greg Bentley; Chief Executive Officer, Nicholas Cumins; and Chief Financial Officer, Werner Andre.

This webcast includes forward-looking statements made as of May 7, 2026, regarding the future results of operations and financial position, business strategy and plans and objectives for future operations of Bentley Systems Inc. All such statements made in or contained during this webcast other than statements of historical fact are forward-looking statements. This webcast will be available for replay on Bentley Systems Investor Relations website at investors.bentley.com on May 7, 2026. After our presentation, we will conclude with Q&A.

And with that, let me introduce the Executive Chair of Bentley Systems, Greg Bentley.

Thanks to each of you for your interest in BSY's 26 Q1. Nicholas will describe factors that contributed to commendable operating performance, favorably according as usual for BSY with our annual full year outlook. Werner will put this in the financial terms, which continue to differentiate BSY as leading among peers, both in the quality and in the measures most meaningful to shareholders of sustained profitability and cash flows. I will supplement their on-the-ground reporting with perspectives behind our characteristically even higher prioritized endeavors to benefit our future value, in particular, through advancements which make AI for us more a seminal opportunity than a terminal threat.

Last quarter, I enumerated some respects in which Bentley Systems' prospects are rather uniquely enhanced and accelerated by AI. Our advantages as summarized here, make the case that leadership in infrastructure AI is destined given our experience and determination within Bentley Systems' grasp. Our long-established incumbency as the digital quartermaster for infrastructure engineering organizations substantive tooling is a differentiating and immediate advantage. Quantification of that pole position will be my focus today.

In particular, our modeling and simulation applications have established the de-facto standards for responsible and deterministic infrastructure engineering, and the stewardship for each account of their cumulative infrastructure engineering data in Bentley Infrastructure Cloud, ProjectWise, SYNCHRO and AssetWise, positions them to leverage AI to compound value for themselves and for their own constituents at a steeper rate than ever.

Beyond our existing consumption business models, AI is on the cusp of adding to our growth incrementally, agentic API consumption of our modeling and simulation functionality, especially for optimization of designs and, for instance, to intrinsically improve constructability. And our Asset Analytics , spawned by AI and already exceeding $50 million in annual revenue run rate, is leading the way toward instant on digital twins to optimize operations and maintenance, commercialized through subscriptions denominated in consumption per asset.

Imbued with all these factors and shaping our distinctive AI game plan, our business is anchored by stalwart enterprise accounts. Within infrastructure engineering, these enterprises are on the leading edge of adopting, acting upon and evolving individual proprietary AI initiatives, which we are there to support, prioritize and enable. What will never change is that their business is our business, and their success is our success.

Underscoring this affinity, 45% of our revenue comes from 220 accounts, which each spend over $1 million per year with us. And almost 2/3 of our revenue comes from 824 accounts, which each spend over $250,000 per year with us, mostly, of course, through E365 consumption subscriptions, which include, in each case, a dedicated BSY technical success team positioned to nurture joint AI initiatives.

To understand the extent to which our interests are aligned with rather than opposed to our accounts, let us drill down on project delivery firms, in particular, because engineering news record surveys and ranks them all, publishing individual revenue breakdowns that can help us understand their economics and our own penetration level and potential as digital quartermasters.

From engineering news records, 2 lists ranking, respectively, domestic and international headquartered design firms, last published together a year ago, we compiled the composite ENR Global top design firms ranked by their verified design billings. For 2025, these 639 global top design firms generated $280 billion of design billings. 25% of these design billings were generated by 25 Chinese organizations. Unfortunately, because they're generally within state-owned entities, geopolitical tensions currently inhibit their accessibility for us.

Thus, the top global design firms ex China consists of 614 top firms generating design billings of $212 billion. Of these, 470 are BSY accounts, together accounting for $198 billion in design billings, 93% of the ex China total. A considerable portion of those whom are not the BSY accounts are rather pure architecture firms.

With our ARR across these BSY accounts totaling $414 million or about 28% of our overall ARR, top design firms are our largest constituency. Per $1 million of their design billings, they spent on average about $2,000 in BSY ARR. On average, each uses about 4 among BSY's top brands plus several other lesser brands.

The top brand for these accounts, also now BSY's top brand, is ProjectWise, in use by 270 of the top firms, together generating $160 billion in design billings, representing 76% of the design billings of the ex China top firms. In our coming quarter, we will describe joint AI initiatives with these firms to compound the reuse benefits from their Bentley Infrastructure Cloud data platform.

For now, I would like to quantifiably illustrate the aligned incentives for BSY and top design firms to work together on the incipient AI transformation of their business model. The key is that infrastructure engineering software isn't for these firms overhead or administration. It is a most necessary factor of production to enable, capture and deliver their substantive work. Along with attendant labor and associated computing, it's a cost of revenue.

To assess the economics, let's consider a representative project with $1 million of design billings. Triangulating variously, including from BSY's $2,000 average of spending per $1 million of design billings across the universe of all of our top firms accounts projects globally, I estimate that a representative $1 million design project would consume about $10,000 of design software. Since margins for these firms now reach about 10%, the cost other than for design software, mostly for engineers labor, must then be $890,000.

Now what could a top firm gain by theoretically investing to the detriment of its design software vendors to somehow reduce its design software spending, say, by 20%? For that putative inspiration and effort and distraction and risk and liability and investment, which thus couldn't afford to be much, their net profit margin would extensively grow from 10% to 10.2%.

But back to the drawing board, consider that investing instead in AI automation and agentic API consumption and computing, even if that say double the all-in design software cost, could save at least 20% of scarce engineering time as that is readily achievable with just agentic automation of drawings production. It seems logical to prefer reducing engineering labor because these top design firms compete with one another for a demographically shrinking talent pool of infrastructure engineers.

And with near record backlogs, these resource constraints limit the new projects each firm can pursue, even though there's an abundance of infrastructure engineering work available. Unfortunately, with the prevalent hourly billing commercial model for infrastructure engineering, improved productivity would produce more output with the same staffing, but not more design billings. So one more simple change would be needed to make this fully worthwhile, transform the design billings commercial model for such projects to fixed pricing.

Now if I were an infrastructure owner-operator client, the only objection I can think of to fix pricing would be a concern that as a result, corners might be cut, alternatives might not be pursued and thus, engineering quality could be compromised. But now with API consumption enabling demonstrable agentic optimization, that concern can be overcome technically. And each human engineer, now augmented by busy agents and automation, remains daily in the loop, but delivering 20% more design billings on additional projects in the same time, yielding very worthwhile, AI-enhanced economics.

The top design firm enterprise by investing in proprietary AI agents to automate and leverage and to API consume trusted modeling and simulation functionality could now generate IP level margins of over 24% on the same engineering inputs. As the owner client gets better assured and more timely quality of designs at no higher cost, everyone wins, including the fully employed and probably gainfully incented engineer, not to mention the software and computing providers.

So while respecting confidentiality of our enterprise accounts individual AI plans and strategies, Nicholas will provide a brief update on our infrastructure AI program with accounts for such joint initiatives, among his other content.

So over to you, Nicholas. Thank you.

Thank you, Greg. We began 2026 on a strong footing. Our Q1 performance demonstrates our ability to consistently execute against the backdrop of sustained global demand for better performing and more resilient infrastructure.

Before going further, my thoughts are with our colleagues and users affected by the conflict in the Middle East. I am incredibly grateful for our team in the region. Their commitment to our users has been unwavering, and their dedication is inspiring.

Following up on Greg's remarks, I will start with an update on AI. We continue to execute on our AI initiatives across the portfolio, but I will focus on 2 recent highlights. First, on Bentley Asset Analytics. To take this business to the next level of scale, I am delighted to welcome Bryan Friehauf as our new General Manager. Bryan joins us from GE Vernova and was previously at Hitachi. He brings a wealth of experience in scaling enterprise software businesses in the operations and maintenance phase of the infrastructure life cycle.

Second, on Bentley Open Applications. We have engaged with leading infrastructure organizations, both engineering services firms and owner operators, as part of the infrastructure AI initiative announced at the end of 2025. The feedback has been clear and consistent. There is strong demand for us to instrument our applications so they can power users on AI-driven workflows.

Based on this strong feedback, we have prioritized the development of both APIs and MCP servers. In fact, we just released an MCP server for STAAD, our flagship product for structural analysis. To give you a sense of the potential, this allows an AI agent like Claude to interact directly with STAAD to optimize the structural design at machine speed. The ability to iterate on complex design trade-offs so quickly is transformative. Our next steps are to instrument other key applications and to validate the commercial model for this powerful new usage pattern. We look forward to sharing our progress.

Now turning to our business highlights. Our year-over-year ARR growth for Q1 was 11.5%, in line with our expectations. Our net revenue retention rate remained strong at 109%, consistent with previous quarters and highlighting the stability and growth within our existing accounts. Our Enterprise 365 commercial program continues to drive steady growth, both in terms of conversions as well as floor uplift and renewals, noting that Q1 is our smallest quarter for renewals. New logos contributed again, 300 basis points of ARR growth, primarily within the SMB segment.

Through Virtuoso, our flagship commercial program for SMB accounts, we again added over 600 new logos in Q1. At the same time, our growth model is evolving, with an increasing contribution from cross-selling and upselling to existing Virtuoso accounts. While our renewal rate remains high, the sheer scale of the Virtuoso base creates a natural churn dollar amount to overcome each period. Our combination of new logos and existing account expansion allows us to continue to deliver strong net growth.

Turning to our performance by infrastructure sector. Resources was the fastest-growing sector in total and across each geographic region. We expect another strong year for Seequent, bolstered by improving mining fundamentals. I will come back to how we plan to expand our addressable market even further into critical resources in a few minutes.

Our largest sector, Public Works Utilities, delivered a solid quarter, driven by robust global infrastructure investments. Power Line System continues to benefit from increasing demand for grid resiliency and new power generation as well as international expansion. The adoption of Seequent applications for the civil infrastructure also supported growth in that sector. Growth in the industrial sector continued to be solid, while commercial facilities remain relatively flat.

Turning to our tone of business by geographic region. In the Americas, our largest region, the U.S. delivered solid growth, supported by stable public funding at both the federal and state levels for transportation, grid and water projects. Private investment was also robust, particularly in resources and AI-related data centers and power generation.

Latin America delivered a standout quarter, with strong performance from Seequent in mining and from our increased focus on transportation in the region.

EMEA was our fastest-growing region in the quarter. This strong performance was achieved despite the conflict in the Middle East, which saw some project delays and a shift in consumption to all the regions. However, this was more than offset by strength elsewhere.

In the U.K., growth accelerated as major projects moved into the delivery phase just as we anticipated last quarter. We also saw robust growth in Africa, driven by increased spending in mining.

Asia Pacific delivered solid growth, with India once again leading the way, and Australia showing improved momentum. While we continue to navigate persistent headwinds in China, which represents approximately 2% of ARR, the strength across the rest of the region more than compensates.

Now I would like to take a deeper dive into our resources sector. This is a part of our business that has become increasingly significant, and we believe it's important for you to appreciate its journey and its forward-looking potential.

When we acquired Seequent almost 5 years ago, our primary objective was strategic, to integrate their best-in-class software for understanding the subsurface into the world of infrastructure. We knew that a misunderstanding of ground conditions is a primary cause of delays in risk in major infrastructure projects. As a byproduct of that strategic move, we also acquired a sizable and thriving business in the resources sector.

I am pleased to report that the initial strategy has proven effective. Since the acquisition of Seequent, we have grown our subsurface ARR in civil infrastructure by a factor of 4, in part due to successful cross-selling into the existing Bentley accounts. The potential for further growth is significant, as engineering services firms adopt a ground informed design approach, bringing detailed ground investigation in as a foundational step before design begins, much like they already do for above-ground surveying.

At the same time, Seequent has continued its impressive growth in its core mining market. Seequent's growth rate in 2025 accelerated as the geopolitical climate and race to AI increased the focus on critical minerals. We expect these trends to continue in 2026, contributing to another standout year for Seequent. However, it is important to note that Seequent has delivered strong growth even during the mining exploration slowdown starting early 2023, production mining companies use our solutions to mine existing deposits more efficiently. But the potential of Seequent in resources extends beyond traditional mining. Seequent's technology is pivotal for other critical resources that are essential to the global economy and to society.

Take, for instance, new sources of energy. Seequent software is already instrumental in the operations of more than 60% of the world's high-temperature geothermal electricity generation. Now it's being applied to new enhanced geothermal systems by companies such as Fervo Energy, a winner at our 2025 Year in Infrastructure Awards. Their project Cape in Utah, for example, demonstrated how new drilling techniques and digital technologies are making geothermal power increasingly accessible and economically viable.

Clean, renewable and consistent baseload energy is more critical now than ever as AI and data centers demand more power. And Seequent impact extends to the most vital resource of all, water. Groundwater supplies about 50% of global domestic water and over 40% of irrigation water. Most of these resources are under stress due to overuse. Seequent software is used around the world by engineering consultancies to manage these resources, mapping aquifers from California to India, designing managed aquifer recharge facilities and constructing a digital framework for groundwater management models. So nearly 5 years since the acquisition of Seequent, resources has become our second largest sector, accounting for more than 20% of our sector attributable ARR, and it continues to be our fastest-growing sector.

In summary, it was a strong start to the year. We are executing well in a robust market, and we're excited about expanding our reach further within the resources sector. Before handing off to Werner, a quick update on our event strategy. We are decoupling our YII Awards from our User Conference to create 2 distinct world-class events. Our Year in Infrastructure event will now be exclusively focused on our global awards competition. We are keeping its intimate and celebratory format. And this year, it will be held in Singapore from October 6 to 7. Separately, we're launching a new large-scale User Conference dedicated to product learning, best practices and community networking. The very first will take place in Toronto in April of 2027. We believe this new format will allow both events to thrive.

And now for a detailed review of our financial results, over to you, Werner.

Thank you, Nicholas.

We are pleased with our strong start to the year, which continues the momentum we built through 2025 and positions us well within our financial outlook for 2026.

Total revenues for the first quarter were $424 million, growing 14.5% year-over-year and 11.9% in constant currency. Our growth continues to be driven by our mainstay subscription revenues, which represent 93% of total revenues and grew 14.7% for the quarter or 12.2% in constant currency.

Our E365 and SMB initiatives continue to be solid contributors. In our smaller and less predictable revenue streams, perpetual license revenues decreased 18% in constant currency. Perpetual license sales remain a very small part of our business and are, as always, a less controllable and less predictable component of our revenue mix. Service revenues increased 25.8% in constant currency, driven by long-awaited reacceleration in Maximo-related services from our cohesive business.

Last 12 months recurring revenues now stand at $1.440 billion, increased by 13.3% year-over-year or 11.5% in constant currency and represent 93% of total revenues. Our last 12 months constant currency account retention rate remained consistent at 99%. Our constant currency net revenue retention rate remained at 109%, consistent with recent quarters. The combination of our high retention rates and new business momentum gives us confidence in the continued durability of our recurring revenue growth.

Now turning to ARR. We ended the first quarter with ARR of $1.495 billion at quarter end spot rates. On a constant currency basis, our year-over-year ARR growth rate was 11.5%. Our sequential quarterly growth rate was 2.5%, all organic and in line with our expectations for the quarter. We continue to expect our quarter-over-quarter ARR growth seasonality to be similar to 2025 and thus organic year-over-year ARR growth rates to be relatively stable during the year. Our GAAP operating income was $126 million for the first quarter. As I've discussed previously, our GAAP results can be impacted by deferred compensation plan revaluations and acquisition-related expenses.

Moving to our primary profitability measure, adjusted operating income less operating stock-based compensation or AOI less operating SBC. As we announced with our 2026 outlook, this is the first quarter we are applying this refined metric. This change aligns the treatment of cash-settled and equity-settled acquisition-related stay bonuses, removing M&A-related volatility from this key operational metric. AOI less operating SBC was $141 million for the quarter with a margin of 33.2%. This quarterly margin performance was in line with our expectations, positioning us well to deliver on our annual margin improvement. As a reminder, we plan to invest earlier this year, resulting in operating expenses being more weighted towards the first half compared to 2025.

Our free cash flow for the quarter was $188 million. This result was in line with our expectations and reflects 2 key factors we signaled on our last earnings call. First, our 2025 free cash flow benefited from exceptionally strong collections at year-end. This created a timing benefit in 2025, which, as anticipated, resulted in a tougher year-over-year comparison for the first quarter. Second, the trend for operating expenses to be relatively more weighted towards the first half this year is reflected in comparison to 2025 for both profitability and cash flows.

Looking beyond quarterly timing fluctuations, on a last 12 months basis, free cash flow of $492 million was up 13%, and we remain on track to meet our full year free cash flow outlook of $500 million to $570 million. We continue to execute a disciplined and balanced approach to capital allocation. During the quarter, we repaid at maturity the outstanding balance of $678 million of our 2026 convertible notes, utilizing borrowings under our credit facility and cash on hand.

The retirement reduced our fully diluted share count by approximately 10.6 million or 3%. In total, our net debt decreased by $134 million in the quarter. We also returned capital to shareholders, deploying $54 million for share repurchases and $21 million for dividends.

Our balance sheet provides significant strategic flexibility. At quarter end, capacity under our credit facility was $756 million, and we reduced our net debt leverage during the quarter from 2.1x to 1.9x adjusted EBITDA. Subsequent to quarter end, we closed on a new $550 million term loan A under the accordion feature of our credit facility. This transaction was completed at attractive terms and used to repay outstanding borrowings under our revolver and lowering our interest cost.

With the term loan in place, total borrowing capacity under our credit facility increased to $1.850 billion. This provides ample capacity to support our strategic priorities, positioning us well ahead of the 2027 notes maturity, while also funding potential programmatic acquisitions, ongoing share repurchases and dividends. We continue to actively manage our interest rate exposure. Our safeguards include the low fixed coupon on our remaining convertible notes and our $200 million interest rate swap expiring in 2030.

And finally, we remain comfortable with our 2026 financial outlook range that we provided just over 2 months ago on our Q4 call. With regards to foreign exchange rates, for the first quarter, the U.S. dollar has strengthened slightly relative to the exchange rates assumed in our 2026 annual financial outlook, resulting in approximately $2 million less revenues from currency changes. If end of April exchange rates would prevail throughout the remainder of the year, our Q2 to Q4 GAAP revenues would be negatively impacted by approximately $3 million relative to the exchange rates assumed in our 2026 financial outlook.

And with that, over to Eric for Q&A. Thank you.

Thanks, Werner. [Operator Instructions] First question comes from Joe Vruwink from Robert Baird.

Great. I guess, Greg, the example you shared is interesting. I think the prospect of supporting $200,000 more in revenue by spending $10,000 on software is something all of your customers would gladly entertain. What do you think about in terms of Bentley's product efforts in terms of bringing that proposition closer to reality? What still needs to be done? And what sort of deliverables do you need to demonstrate to your customers so that they believe and give you the buy-in?

Well, I think the path to that foreseeably is the API consumption. The notion that agents spun up by the engineers can do more iterations in the same time. It not only will deliver a better quality of design, but the engineering firms will be able to substantiate that to the owner operators to accelerate this transformation to fixed pricing.

The -- what's really great about that prospect for us is the more that engineering firms and owner operators are excited about and ready to act on this transformation to a new commercial model. That's all good for our prospects. It's where everyone wins, as I say. It can't be doubted that this will happen with the AI inflection as it is happening in every other service industry. Engineering is one of those. But it's been slow until now. And this notion of optimization that literally is in front of us at the moment is going to speed it up, I think.

Our next question comes from Matt Hedberg from RBC.

Great. Good start to the year. When we look at your business, you guys all highlighted a number of, I think, really interesting company-specific catalysts that seem to be moving in your favor this year. And to us, when we sit back and look, it seems to present an opportunity that could push constant currency ARR towards that higher end of the range this year and perhaps accelerate versus last year.

I guess when I sit back and think about all these opportunities, if you were to highlight the 1 or 2 things that you're like these are the most important things that could deliver those results, what would they be? Because it seems like there's a lot of opportunity here for you guys.

Well, I think you're characterizing it exactly the right way, Matt. There's a lot going on and that is very positive for the company. We benefit from very strong tailwinds here in both infrastructure and in resources and as I highlighted during the prepared remarks. So for us to get to the, let's say, higher part of the range, we would need both resources and mining, in particular, to continue to go strong throughout the year. And right now, everything is signaling that this is trending well.

We need Bentley Asset Analytics to continue to grow strongly throughout the year. We need, of course, the core business infrastructure to continue to go well. And then probably another programmatic acquisition, yes. So if you have all of that lighting up, then yes, we're talking about the higher part of the range.

Next question comes from Jason Celino from KeyBanc.

Great. This actually goes back to Joe's question. There's growing debate among the investment community on whether to charge or not charge for access to data. It seems like Bentley is one of the only software platforms trying to monetize this way through API consumption and frankly, might be the preferred way investors may want to see it.

But do you foresee any risks from like a customer perspective on charging this way since other horizontal software companies have decided to be more open and not necessarily charge for it. Nicholas, the STAAD MCP server that you talked about, are there other MCP servers that you might need to stand up?

Definitely, right? So maybe a point of clarification first. When we're talking about API consumption, we're talking about an indirect usage to our engineering applications. So as opposed to having a user directly interacting with the applications, we have an AI agent that is directly interacting with the application. Now there's still a user behind the AI agent, but it's the AI agent, which is now being able to do with the application, a level of optimization, which was simply not possible for an engineer on its own, okay?

Then we have Bentley Infrastructure Cloud, where the data actually resides. And here, we're not monetizing the access to that data because that data belongs to the infrastructure organization that are entrusting us with that data as part of the platform, right? So those are 2 different things.

The ability to leverage AI to optimize designs or even generate parts of the design is just a fantastic value proposition. Clearly, our user-based pricing right now, whether we're talking about attended consumption as part of E365 or annual subscription with Virtuoso is not going to capture the full value that is going to be created, right?

So we're discussing with the representative accounts who are involved in our Co-Innovation initiative called Infrastructure AI. We're talking about a different commercial approach where it will be indeed an API consumption-based pricing, maybe token-based because that seems to be the common denominator now across different AI tools.

But that's really for the engineering applications. For Bentley Infrastructure Cloud, at least in the foreseeable future, the pricing is indeed user-based, either user-based consumption or attended consumption or annual subscriptions.

And may I make clear that today, we only monetize attended consumption. Our strategy is to introduce and increase the API consumption and give users and accounts a chance to explore the potential of that and learn in the process what it costs us, what it could cost them and what the benefits and values are, so that we can arrive at an appropriate monetization approach to that, which we are open-minded about for now.

Next question comes from Siti Panigrahi from Mizuho.

Great. Just following up that AI part, Greg, you laid out really a compelling case for AI within your customer base. Wondering what sort of feedback have you got from your customer? And how should we think about the TAM expansion for Bentley or even the ARPU uplift potential as AI drives even deeper multiproduct adoption? Maybe it's not near term, but how should we think about ARPU and TAM expansion?

And in the same context, Werner, how are you looking about the margin and even there is an incremental engineering and computing cost implication as Bentley start focusing on building AI products. How should we think about the investment intensity over the next 12 to 18 months?

I'm going to ask Nicholas to speak about the account reaction. But let me just say a particular multiproduct scenario that has me excited is during design to be able to have an agent using SYNCHRO to explore the constructability of what's being designed while it's being designed. This doesn't imply that the design firms that we talked about will be doing the construction necessarily. Someone will be doing the construction subsequent to their involvement, but they'll be able to say to the owner operator, we're going to be able to give you a design where we already know the economics of construction. We've simulated that in the course of our design, which is an example of something not even conceivable now that AI will introduce through API consumption.

But Nicholas, perhaps you can speak to the reaction of accounts to the prospects for this instrumentation.

I'll do that, and then I'll also get to the TAM question.

So the reaction of our accounts is really validating the opportunity for this indirect usage of applications. Now we do see a difference here between the very large infrastructure organization, in particular, the very large engineering services firms and all the others.

The very large ones, by the way, exactly like it was shared with the [ AC Adviser ] CEO survey of 2025. The very large ones are the ones who are really investing in their own AI-driven workflows. They are the ones who are exploring with us how exactly they will start to use our applications indirectly through AI, right? Now all of the infrastructure organizations we talk to, all of them, whether they're big or small, are welcoming that we deliver our own AI capabilities as part of our products, right?

Now back to the very large one. These are still very early days. And as we commented actually last quarter, we're much more in the mode of exploring and validating and then confirming, for example, that, yes, an MCP server for STAAD makes a lot of sense, yes, and then an MCP server for an application to another application. And we're very focused on adoption as well. The monetization will come next, where we are doing monetization right now with AI is really through a different offering, which is Bentley Asset Analytics, which we can talk again in a moment.

But now back to your question about the TAM. I think it's a great question because the TAM that we shared at the moment of our IPO when we became public was all based on the number of engineers. And effectively, if we're talking about indirect usage of our applications through APIs, we are somewhat removing that natural cap of how many engineers are out there and how much can they consume within a given day in terms of applications for Bentley. So we're very excited about that as a long -- let's say, longer-term growth opportunity for us to actually expand our TAM by having indirect usage of application through AI.

And Werner?

Yes, maybe on the margin. So I think as Nicholas said, it's early days. It's completely immaterial as of now. But I could see that gross margins will be impacted, but it's likely, and this has to be considered then in the monetization of the products.

The next question comes from Dylan Becker from William Blair.

Appreciate it. Maybe Greg or Nicholas here, I think the shift to outcomes obviously makes sense and kind of your ability to facilitate that efficiency gain, I think, is abundantly clear. But I was maybe wondering on the importance of kind of your services or in the direct customer relationships you have from a go-to-market perspective, and how you can kind of help those customers bridge that gap from a change management perspective and kind of help maybe accelerate that push in the economic delivery model, if that makes sense.

One thing that I'll remark upon is that our dedicated teams in E365 over the past year or 18 months have discovered a new productive way to help our accounts. It's by helping them pursue new business. It's helping join their pursuit teams to bring new ideas to owner operators. And I hope that will include as we implement the optimization approaches, how that gets sold and communicated among things to enable fixed price, which is the beginning of outcome-based contracting.

Yes. At the end of the day, when it comes to the commercial model of the engineering services firms we serve, this is very much to be decided between them and their clients. And then depending on the industry, it will vary quite a lot. Some of them have already embraced fixed-based pricing. Others are still in time and material.

What we can do, besides, of course, demonstrating the power of AI and how it changes the whole value proposition. Besides that, we have a lot of high-level advocacy efforts that we're doing with governments, with the financial sector as well, with the clients. Some of them are through our infrastructure policy advancement team. Some of it is also through our services arm called Cohesive, which is providing consulting to owner operators and gaining insights about what else could be done, what's the art of the possible and potentially what does it mean for their commercial model.

When Nicholas describes some have begun fixed pricing, it's primarily those that do private sector work, where private sector owner operators are quicker to adapt and evolve than government-funded public works and utilities.

The next question comes from Jay Vleeschhouwer from Griffin.

Nicholas, a quarter ago on the call, you described Bentley Infrastructure Cloud as your data foundation for AI. And I assume it's more broadly so aside from AI that in terms of the importance of infrastructure cloud. With that in mind, could you talk about perhaps some of the adoption metrics for infrastructure cloud? Do you think of it as a kind of prerequisite for or leading indicator for other business, AI or not?

And then if I may, a point of clarification, given all the API discussion earlier, in a complex federated system of that kind, how do you assure when customers adopted good or better throughput or performance in a complex API system?

So to the first question, when we're referring to Bentley Infrastructure Cloud as a data foundation for AI, we're really talking about the AI efforts of the infrastructure organizations we serve. So when they upload files, when they connect data systems into Bentley Infrastructure Cloud and all of the data from those files or the systems is mapped to a Schema so that that data becomes not only queryable, it can be analyzed, but it can also be leveraged by AI, right?

Now we made it very clear that we are not using that data, which is in Bentley Infrastructure Cloud to train our own AI. That definitely sets us apart from other software providers out there. Now data ownership is a very, very sensitive topic across all industries for sure, in infrastructure as well. We do not think it's right to be leveraging the intellectual property of some companies to train AI that will benefit other companies, right? We just don't think it's right.

So we're not going to do it. So the only time we use data within that Infrastructure Cloud to train AI is when we're explicitly asked by infrastructure organization to do so. And then it really is their data. And by the way, we get full transparency of what data has been used in order to train AI. That sets us apart to the extent that it really becomes a reason why infrastructure organizations also choose Bentley Infrastructure Cloud as opposed to other benefits because they really have trust that we're going to be a good custodian of that data on their behalf.

Now when it comes to API throughput, indeed a point of clarification. So for example, with the demo I was showing with STAAD, we are, in a sense, in the first step of the instrumentation where these applications are still running on the desktop. In fact, you can see it with the video, it was still running on the desktop. They're not yet platform services. That will be the next step.

So the first thing we're doing is making sure that the APIs exist, that there are MCP service available. So it's easy for AI agents to interact with those applications. But all of that interaction is actually happening on the local computing station on the personal computer. When -- at some point, of course, that's what we're discussing also with accounts, we're talking about the longer term here. Some of those workloads will be moved to the cloud to become true cloud services, but then we'll just make sure they are as performant as they should be like we do for all of the cloud services that we offer.

The next question comes from Alexei Gogolev from JPMorgan.

Great to see you all. You've mentioned that it is still early days for applying AI to mission-critical engineering and you are leading the exploration to building and drive the adoption of highest value AI-powered workflow. So could you maybe speak about some of the initiatives on that end, what efforts you're taking?

So we have -- if you're referring also to one of the earlier comments around we're making progress with our initiatives across the portfolio. We have both efforts to deliver our own AI capabilities as part of our products and then efforts to instrument our applications, in particular, the engineering applications so that they can support AI-driven workflows that are being created by the infrastructure organizations that we serve, right? So those are 2.

For the latter one, there is a Co-innovation initiative that we've launched at our annual conference last year, YII, so towards the end of 2025, where we invited infrastructure organization to join, talk to us and explore with us what are the use cases they're going -- they would like us to be able to support in their applications. And it was no surprise that the actually first use case that identified was for STAAD because we had already seen STAAD, which is our flagship product for structural analysis, being used this way because it had an API in some of the submissions for the Going Digital Awards at our annual conference. Over multiple years, by the way, it was clearly the first one. So no surprise that there was a very strong demand for us to create an MCP server based on that API in order to support the interaction of AI agents with that.

And so that's the way we're interacting with them in that context, which is understanding what are the use cases with the biggest potential, validating with them and then creating those MCP servers. And we have multiple efforts across our pretty wide portfolio of engineering applications to create the MCP servers directly if the APIs already exist, if not start to create the APIs themselves, okay?

But then now for the AI capabilities that we offer, we have AI capabilities as part of the engineering applications that will automate parts of the design workflow. Typically, we go for the, let's say, Mundane tasks, which are extremely time-consuming like drawing production that we talked about in multiple quarters ago.

We have AI capabilities built within Bentley Infrastructure Cloud, for example, in order to facilitate the search of engineering data using natural language. We have AI capabilities being part of Bentley Infrastructure Cloud in order to help navigate very complex construction models as part of SYNCHRO. We obviously have AI pretty much for all of our applications for Bentley Asset Analytics, whether we've developed them or we acquire them. And we even have AI in Seequent, which I spent time highlighting today as part of my prepared remarks, right? So there are efforts related to AI busy really across the portfolio. It's a wide range of initiatives.

The next question comes from Daniel Jester from BMO.

This is Will Hancock on for Dan Jester. So you hit on data ownership a few minutes ago. And I just wanted to double-click there. Are you guys seeing any shifts in customer behavior, whether that be increased willingness to use data to train AI models or conversely more cautions around permissions?

And then are you guys seeing any variance across end market or customer size?

So what hasn't changed is the fact that it's the -- let's say, the largest infrastructure organization, especially the largest engine services firms that are the most advanced in their own AI efforts and therefore, looking into their own data on how they can train their own AI models. So that hasn't changed.

What we have seen in the last few months is a very suddenly infrastructure organization realizing that not all software providers have a principal approach to data. And then therefore, pushing for a lot of clarification on terms and conditions and access right, et cetera, for the data. That's what really sets us apart, which is already back in 2023, we took this very principled approach. Our commitment to data stewardship that made it very clear that our users' data is their data always. And we don't use it to train our AI unless explicitly directed to do so by the infrastructure organizations we serve.

It is indeed a very sensitive topic, and I expect it as Bentley infrastructure organization become more and more educated about on one hand, the potential of AI. And second, the fact that there is a difference here across software providers, I do expect this topic to become even more sensitive going forward.

And I'll add that while the immediate attention is paid to training AI models. The -- I think the ultimate value of this data will be its reuse in future designs. That's never been the norm because in attended consumption, an engineer will always prefer to start with a blank screen and originate a new approach. But each firm has a very valuable archive of project data in ProjectWise Bentley Infrastructure Cloud and AI will be good at finding and suggesting and modularizing and parameterizing.

And ultimately, when that reuse can be informed by the operating and maintenance performance of those designs, which the engineering firms will increasingly be in the business of improving and optimizing. That will be a virtuous cycle that will yet reinforce the valuable proprietary advantage. When I say that engineering firms can earn IP returns, I mean returns on that valuable data and knowledge, their accumulated compounding advantage with Bentley Infrastructure Cloud.

And maybe I'll put one final point because it is such an important topic because some of you may ask, but wait a minute, are the other software providers who are shamelessly leveraging the data that is stored in their platform to train their AI, do they have an advantage? And I will clearly say no. And for 2 reasons. Number one is because of their stance, infrastructure organizations have a higher tendency of choosing us and our platform because they can trust us. And second is for all of the AI capabilities that we're developing that I mentioned before as part of our initiatives across the portfolio, right?

We don't do it in isolation. We do it always with representative accounts. And we've never had an issue of not having enough data to train our AI models, either by using our own synthetic data or getting data from those representative accounts involved that they deemed they think is not sensitive, not differentiating and they're happy to contribute, right? So we clearly see this as a net positive for -- and a net advantage for us.

The next question comes from Taylor McGinnis from UBS.

Yes. Can you guys hear me?

Yes.

Yes.

Okay. Perfect. So if I look at your constant currency net new ARR growth historically, I think it's been around $26 million to $27 million. And this quarter, it was $37 million, so up 36% and really strong. So when you just think about the drivers of what drove that strength in the quarter, could you unpack a little bit of that? And was there anything onetime or different seasonally this year compared to what you've seen in the past that might explain some of that? And then just how do we think about that in the context of what you're thinking about demand trends going forward?

So maybe I take that. So Q1 saw a continued strong momentum as we exited 2025. And the quarter-over-quarter growth puts us somewhat ahead of what we saw in Q1 2025. It was, I would say, predominantly led kind of the -- over by the continued momentum of our Seequent business and mining just doing still very well. So as we said, that puts us in a pretty good position for our full year outlook range. But it's still -- Q1 is 20% of our ARR opportunity based on contract resets. And it's -- got to see that the rest of the year plays out well as well.

The final question comes from Joshua Tilton from Wolfe Research.

I'll give you guys a break from the AI questions. Maybe just 2 points of clarification on my end. First, you talked about Seequent being really strong last year and expectations for it to be strong again this year. Are you -- do you expect it to accelerate again in 2026?

And then my second part to that question is, on the other side of that, you kind of cautioned us -- or maybe I'm reading into it a little bit, but you cautioned us around Virtuoso getting big and a churn dollar amount that you have to overcome each quarter. Is there anything we should think about or be paying attention to or anything unusual about that churn that you're trying to signal to us this quarter?

Okay. I'll try to be brief since we are at the end of the scheduled time. So on Seequent, no, we don't expect further acceleration in a sense that as part of our plan for 2026, we assume the same level of growth that we've seen towards the end of 2025, right? So clearly, in 2025, there was an acceleration. And for 2026, we just assume this is going to continue and Q1 basically proves us correct, yes?

On Virtuoso now, we just wanted to explain that in addition to solid growth coming from new logos, we now also have growth coming from existing accounts because in previous earnings calls, we were only talking about new logos. Now clearly, we've developed a new muscle here, which, by the way, also helps for retention. There's a clear correlation, which is if we see accounts using more than one product, then there will have a higher propensity of staying with us.

But the overall retention rate remains stable at a high double digit, very similar to what we've shared in previous quarters. It's just mathematics. It's a much higher base, of course. So indeed, there is a churn amount in terms of dollars that needs to be compensated for. But that's it. It was just explaining all the different puts and takes and explains why even though we have this new muscle, you can see the growth being still very consistent with Virtuoso.

Maybe I'll just say in closing, your first question about acceleration. There hasn't been questions, particularly about geopolitical events in the world. But the notion that each country needs to be more self-sufficient in its defense and its resources and so forth represents a commitment to investment in infrastructure, and it's not limited to resources. You need the resources for infrastructure. You need the infrastructure for resources. And I expect that to be not a short-term phenomenon to our benefit.

Thanks. That concludes our call today. We thank you for your interest and time in Bentley Systems, and we'll talk to you again next quarter.

Thank you.

Good morning. Thank you for joining Bentley Systems Q4 and Full Year 2025 Results and 2026 financial outlook. I'm Eric Boyer, Bentley Systems Investor Relations Officer.

On the webcast today, we have Bentley Systems Executive Chair, Greg Bentley; Chief Executive Officer, Nicholas Cumins; and Chief Financial Officer, Bernard Andre.

This webcast includes forward-looking statements made as of February 26, 2026 regarding the future results of operations and financial position, business strategy and objectives for future operations of Bentley Systems Inc. All such statements made in or contained during this webcast other than statements of historical fact are forward-looking statements. This webcast will be available for replay of Bentley Systems Investor Relations website at investors.bentley.com on February 26, 2016.

After our presentation, we'll conclude with Q&A. And with that, let me introduce the Executive Chair of Bentley Systems, Greg Bentley.

Thanks to each of you for your interest and attention. Nicholas will review the factors behind our STAR 2025 operating results, and Werner will then provide our consistent 2026 outlook. So I will take a longer-term perspective.

First, as I hope we'll be welcoming some new investors, I'll begin with an update on Bentley Systems' financial fundamentals. With 2025 results, I can extend my previous review of our first 5 years as a public company to now focus on the years since pandemic disruptions. Our business model prioritizes durability and visibility.

Our key metric, annual constant currency ARR growth has been reliably sustained in the low double digits since 2022. This business performance measure excludes the ARR onboarded with our major platform acquisitions, Seequent and Power line systems. In 2025, the portion of this growth from smaller programmatic acquisitions was at a low of under 40 basis points. During the earlier of these years, rates of inflation as shown here for the U.S. were significantly higher. So in ex inflation real terms, our ARR growth has been more than maintained.

Reflecting our standing commitment to annually improve profitability as we gain efficiencies, especially from our 94% direct sales motion, our adjusted operating income grows faster than revenue. As we regard stock-based compensation as fungible with cash, our key profit measure is AOI less SBC. Having institutionalized annual improvement of about 100 basis points in margin since 2022, we have compounded AOS less SBC dollars at over 16% per year. Our straightforward revenue recognition primarily ratable and almost never for multiple years and annually prepaid subscriptions make free cash flow also predictable though subject to variations in working capital, taxes and interest.

With such factors having been favorable in 2025, our free cash flow margin reached 35%. However, cash flow isn't truly free to the extent it must be allocated to offset share dilution from stock-based compensation. Our truly free cash flow margin, that is less SBC, reached 30% in 2025. For valuation benchmarking, one must reckon per share. Our fully diluted share count has been substantially constant including putative dilution from the convertible debt that funded our platform acquisitions. But last month, we retired the maturing 2026 convertible debt as will presumably occur again next year, reducing our fully diluted share count by about 3%. And having reached a satisfactory target range of about 2x, we have completed delevering since the platform acquisitions and can now allocate more of our cash flow generation towards scaled up programmatic acquisitions.

The consistent 2026 financial outlook Werner will present, reflects our confidence in both our robust end markets and in sustaining our execution fundamentals. But market perceived risks of AI interloping seem to have discounted our value thereafter to nearly terminal. In fact, for Bentley Systems, AI is not a risk to be countered but an unprecedented opportunity. Distinctive fundamentals of infrastructure engineering serve substantially in our favor. As the industry's established and trusted digital quarter master we are best positioned to catalyze with infrastructure engineering organizations, the value to be realized by taking full advantage of AI's potential to transform the substance of their work.

Over 42 years, our key advantage has been providing continuity across technology generations, something highly valued for long-lived infrastructure projects, assets and engineering careers. Based on our actual experience over these decades, enabling and then encouraging progress from CAD to BIM to digital twins, the faster AI and its integration improves the better for Bentley Systems. The deliberate pace of technology adoption in infrastructure and engineering is rooted in legitimate prudence.

Each of our lives and much of their quality depends on vital infrastructure, meeting standards for safety, resilience and fitness for preps. This is why specific to public infrastructure, regulatory regimes variously require a licensed professional engineer to personally seal project deliverables, vouching under penalty of law and of liability that they supervise the work. This requirement cannot be met by casually adopted unproven AI tools.

Institutionally and contractually, project collaborators across engineering disciplines, must adhere to formally structured interactions and data formats. Owner operators and engineering enterprises mandate strictly approved tool sets for interoperability and quality assurance. In this world, do-it-yourself AI tools without years of vetting would confine and engineer to trivial work at their own risk.

From a practical standpoint, the nature of our applications is unlike the administrative software now suspected to be vulnerable to AI replacement. Like other professionals, infrastructure engineers do use administrative software but not from Bentley systems. Our applications are virtually devoid of the forms, transactions and text that characterize administrative work. The screen capture as you see here, include, by the way, a data center site as typically construction modeled by DPR construction, a world leader in virtual design and construction.

Infrastructure engineering is performed through immersive, interactive, 3D geospatial modeling experiences like these with almost all projects juxtaposed within real-world brownfield environments. Their design requires all context all the time while orchestrating complex algorithms and simulations. And while engineering is a creative profession, unlike other 3D creators, an engineer cannot be satisfied with the notional abstractions of mirror visualization, often what can't be seen is most important.

Precision is paramount with 0 tolerance for approximation let alone hallucination. Beyond the confidentiality required for physical and cybersecurity of essential infrastructure their designs constitute the valuable intellectual property of engineering firms and their owner-operator clients as by far the long-standing primary system of record for infrastructure design, all data managed through project-wise within Bentley Infrastructure Cloud is strictly proprietary to the engineering organization. We responsibly steward this complex engineering data for their authorized use only, including for AI training, there is no such credential data publicly accessible to be script for such training.

In any case, our users' economic incentives to seek alternatives are perhaps surprisingly mild. Though a mission critical to produce, capture and deliver an engineers work, our software costs on average per user day, only about 3% of that user's burdened daily labor cost. This low substitution rate of technology for labor compared to other industries, is likely rooted in owner-operators archaic norm of paying by the hour and often based on low bid for engineering services.

Perversely, this incenting advances in productivity. Spurred by now chronic shortages of engineering capacity, I believe that AI is poised to transform infrastructure engineering business models to finally compensate not for man-hour inputs, but for better quality outcomes. All votes will be raised, but especially software and computing spending per engineer with AI agents automating design optimization. Engineers could, of course, improve their designs without automation to the extent they would be allowed time to explore more iterations. But with the current technical norm of attended consumption, compounded by the current commercial norm of hourly billing budgets rarely afford such repetitions. AI can break through this bottleneck by enabling an engineer's AI agents to automate the workflow of systematically permuting the engineers initial design over a many-dimensional solution space, for instance, varying, geometry, dimensions, materials, capacities, utilization and so forth.

As a start for this, we are already providing Copilot AI for users to create from natural language, scripts that run against the APIs of some of our applications. Through many more APIs to be instrumented across our portfolio, these automation agents we'll endlessly invoke our proven modeling and simulation functionality and a heuristic search strategy to converge to qualified superior alternatives for the human-in-the-loop to subjectively assess.

But consider that AI could extend design optimization even further. For example, to reuse proven components from past projects and to minimize construction effort, schedule and risk. The potential incremental value of such optimization can reach a very significant portion of infrastructure projects, total installed cost, which together is literally in the trillions of dollars, annually.

Owner operators will willingly pay more for designs accordingly AI optimized for quality. Project delivery teams finally will be able to expand beyond the current constraints of engineers' time and will compete to generate value by leveraging their IP in AI agents and in proprietary project in asset data. For our part, Bentley Systems will, in due course, incrementally monetize API consumption on a scale -- orders of magnitude greater than that of continuing attendant consumption. But a quite immediate opportunity already open for us is to apply AI and digital twins toward optimizing the operations and maintenance life cycle of assets. This is a committed priority of Bentley Systems' new management generation.

Our comparatively small proportion of revenue from asset performance to date in relation to years of investment shows how slow this had been to significantly grow. But more recently, in conjunction with fast improving reality capture technologies, ranging from drones through dash cams AI has enabled instant on digital twins. Our asset analytics strategy accelerated by year-end acquisitions reached the $50 million run rate milestone for asset consumption revenue in 2021.

Our progenitor OpenTower iQ continues its leadership with ARR now in 8 figures. It exemplifies our winning strategy, uniquely combining market-leading digital twin creation with best-in-class engineering simulation. Blyncsy for roadway operations also had a breakthrough 2025 and is now being piloted by many departments of transportation.

Hawaii announced a statewide commitment including providing dash cams to drivers to extend coverage. Alabama is using Blyncsy to improve decisions on maintenance and capital project spending. Our 2 acquisitions were strategically complementary. We acquired the assets of Pointiva, whose R&D and valuable patent portfolio, extend our asset analytics platform in new directions. Pointivo software has been broadly applied for advanced AI-based point cloud processing, automated measurement and condition analysis and inspection workflows.

And we acquired Talon analytics. The leader in asset analytics for telecom and utilities, with capabilities proven at the level of 8-figure contracts. Talon originated drone capture for wireless structures and evolved its AI-based software to expedite construction completions and ongoing maintenance. Talon already relied on our iTwin capture for engineering-grade digital twins. Recently, they collaborated with integrated grid utilities to pioneer AI-based digital twins for electrical distribution poles to automate the structural analysis for hundreds of thousands of poles we help talent to implement API consumption of our SITEOPS software to gauge the potential for this. The U.S. alone has 180 million distribution poles. And for each digital twin inspection and simulation, our Talon asset consumption revenue is in low double digits. By law, distribution pole inspections are required only every 5 years, which is a reason that classification as ARR is not obvious. The best outcome is for digital twins and AI to make annual monitoring affordable and effective.

The next opportunity for our expanded asset analytics platform is to leverage power line systems simulation to improve resilience of electrical transmission tower capacity. That's the infrastructure most needed for AI computing to grow. For API consumption, we are now prioritizing propagation. But I'll consider that the progression we've just talked about from Talon 7-figure API usage to our mid-8 figures of asset consumption revenue to be representative of our potential to monetize AI at scale.

By virtue of Bentley Systems majority family ownership, our compass has always been set to benefit the long term. Our solid financial fundamentals and are directly relevant organizational experience equip us to tolerably bear the marginal risks and volatility inevitably associated with these increased AI investments and ambitions. My assessment is that AI transformation for infrastructure engineering augurs better times than ever for Bentley Systems. Here's to 2026 and beyond.

And now over to Nicholas.

Thank you, Greg. Building on the context you've provided for AI, I want to start today by outlining our strategy, the significant progress we made in 2025 and how we plan to execute on it going forward. Our approach is twofold. We are not only embedding our own AI capabilities into our products but also instrumenting our platforms. So users and partners can build their own AI-driven workflows. We're investing in AI across our entire portfolio. But for this conversation, I want to focus on 3 key areas that are central to our business and represent a comprehensive and principal approach to infrastructure AI.

First, in Bentley open applications, we're leveraging AI to enhance the work of engineers. This includes leveraging AI to automate interactions with our applications such as the python assistant for MicroStation, automate time-consuming design tasks like generating drawings annotations for OpenRoads and even optimize entire designs as seen with the site layout optimization in OpenSite+.

Just as importantly, our applications serve as a critical environment where AI recommendations are not just tested but continuously optimized. This is not a single pass fail gate. It is an iterative process where software is used continuously improve the AI orchestrated design, ensuring it becomes progressively more sound. This process naturally drives greater consumption of our applications core engineering capabilities as they become central to the AI-driven design workflow.

Second, Bentley Asset Analytics, which Greg spoke about, uses a 2-step process. It leverages AI primarily competitive vision to process imagery and detect features on an asset. It then uses bend open applications to understand what those features mean from an engineering perspective. For example, can a tire safety take on more load. The output is actionable engineering intelligence that an AI workflow can then use, for example, to automatically trigger remediation work in a third-party EAM system, like IBM Maximo.

Third, Bentley Infrastructure Cloud serves as the data foundation for AI. This is where the world's leading engineering firms manage the design files for their current and past projects primarily using project wise. PAUSE Our iTwin technology provides the capability to access data from countless file formats and systems and map it to our base infrastructure schema, making it ready for AI. This unlocks tremendous potential. It allows engineers to search past project data using actual language. It will enable our users to fine-tune our AI models with their own proprietary data or even train entirely new customer models. We envision in not-so-distant future where Bentley copilots drawing on an organization's past projects stored in Bentley Infrastructure Cloud can proactively recommend the best design components.

And crucially, as Bentley Infrastructure Cloud maintains a digital thread through operations it will be able to surface invaluable performance data from the field. This will allow users to understand how desires created before have held up over time, providing a historical evidence-based foundation to inform and derisk the next generation of designs. But leveraging historical data is only the first step. The true power of our integrated platforms comes to life when Bentley copilot itself becomes an optimizer. Imagine it not only recommending that proven component, but then offering to reduce its carbon footprint using it in capabilities, refine its foundation with PLAXIS, optimize its structure with STAAD and ensure its constructability with SYNCRO.

That is the unique compiling power of a truly integrated platform for infrastructure AI.

Finally, all of this is built on a principal approach to data. Data ownership is a critical topic in infrastructure. What makes us distinct is our unwavering commitment to data stewardship a principle we first articulated in 2023. Our users and only or users decide if and when their data is used to train AI models.

To enforce this, we provide a data agreement registry, where an account can formally correct or revoke its consent to have its data included in AI training sets, ensuring they remain in full control. I want to be clear about our commercial approach to AI as it is intentionally different across the portfolio. With Bentley Asset analytics, AI is applied to mature operational needs where it delivers tangible ROI.

Our focus there is rightly on driving revenue growth. For the foundational area of AI in design, however, we are playing a longer game. These are still early days for applying AI to mission-critical engineering. Therefore, our immediate priority is to lead the exploration for the highest value AI power workflows while at the same time, building the market and driving their adoption rather than focusing on direct monetization.

As the infrastructure engineering software company, we believe it's our responsibility to lead this transition thoughtfully. This means actively engaging across the entire infrastructure ecosystem. From our deep collaborations with engineering firms and owner operators, to policy discussions with government bodies and partnerships with other technology leaders. We are helping to establish the standards and trust necessary for this technology to be adopted safely and effectively.

Through these efforts, we are building the foundation of usage and trust first, confident that monetization will follow as these new AI powered workflows mature and improve their immense value. So that is a strategic foundation we are building for our next phase of growth.

Now turning to our results from the fourth quarter. Our performance shows the continued strength of our established business today. We delivered a strong finish to the year, and that momentum gives us confidence in our 2026 outlook, where we aim to deliver another year of compounding results within our financial framework. I want to thank all of our colleagues for their dedication and hard work and our users for their continued trust.

Q4 ARR increased 11.5% year-over-year, which was a solid increase from Q3 as we expected. Net revenue retention was stable at 109%. The E365 performance remained steady, and we added 300 basis points of ARR growth from new logos once again, primarily within the SME segment. For the 16th consecutive quarter, we added at least 600 new SME logos through our line store, with retention in this segment remaining high.

Turning to our total business by infrastructure sector. Resources was once again our fastest-growing sector. The standard growth of Seequent is expanding our addressable market into critical resources, a domain that includes mining as well as new energy sources and groundwater. The performance of Seequent, even during the recent mining slowdown, demonstrates the resilience of its business model, which is deeply embedded in operational workflows rather than cyclical capital projects. As market conditions in mining continue to improve, we are confident Seequent will remain a key growth engine for us in 2026 and beyond.

Our largest sector, public works utilities delivered another quarter of strong growth, driven by sustained global infrastructure investment and the standout performance of Bentley Asset Analytics. Power Line Systems also continues to be a key growth driver, benefiting from global demand for grid resilience and increased power generation.

Growth in the industrial sector was solid, while commercial facilities remain relatively flat.

Turning to our total business by geographic region. Our largest region, the Americas, saw another quarter of strong growth. This was driven by a favorable macro backdrop for infrastructure investment that we expect to continue into 2026. The U.S. market remains healthy with stable public funding, ensuring that project backlogs remain large, and engineering services firms busy. We also see our accounts already at capacity, capitalizing on the large private sector investments in data centers, driven primarily by demand for AI computes.

Even though the core design work is more akin to building design, the immense strain these projects place on the local infrastructure, in particular, the power grade and water network creates a sustained tailwind for a broad portfolio of infrastructure engineering applications. Beyond infrastructure, the U.S. administration's announcement of Project [indiscernible] a $12 billion investment to establish a strategic reserve for critical minerals is a clear signal of a broader global trend, the imperative to secure domestic resources, which in turn drives the mining activity that benefits Seequent.

Growth in EMEA was once again led by the Middle East. We expect this exceptional performance to continue in 2026. As investments there shift towards transportation, utilities and mining, playing to our strength even more. Europe delivered a strong quarter with infrastructure clearly remaining a top priority for the EU. This was evidenced by several policy initiatives published in Q4, targeting key strategic goals energy transition, transport connectivity and supply chain security. The U.K. was softer in Q4, reflecting the tail end of project positive from earlier in 2025.

However, looking ahead into 2026, the pipeline for design and energy work is improving significantly. For instance, we were very encouraged by last month's green light for Northern Powerhouse Rail. This adds another multibillion pound project to the design pipeline alongside massive engineering efforts now underway such as size wells.

In Asia Pacific, India delivered solid growth. The long-term outlook here remains strong. In addition to ongoing investments in transportation and the world infrastructure, the country's 2047 vision calls for a massive investment in grid modernization to provide power for all. With Paradise Systems as the industry standard for transmission engineering, we are uniquely positioned to help.

China, which represents approximately 2% of ARR continue to be impacted in the quarter by the economic and geopolitical headwinds, which are likely to remain through 2026. Growth in Australia is showing signs of recovery as headwinds from government changes and a pause in transportation projects subside. We expect stronger growth in 2026 driven by a resurgence in the mining sector and new infrastructure projects related to the Olympics.

All in all, we are very pleased with the continued strength of our business, and we are well positioned with a great foundation and strategy to help the infrastructure ecosystem, leverage AI to deliver even better outcomes. And now for a detailed review of our financial results and outlook for 2026. Over to you, Werner.

Thank you, Nicholas. We are pleased with our finish to a solid year of financial performance, which marks a strong close to 2025. We delivered strong financial results for both the fourth quarter and the full year. For the full year 2025, total revenues were $1.52 billion, growing 11% on a reported basis and 10% in constant currency. For the fourth quarter, total revenues were $392 million, an increase of 12% reported and 10% in constant currency. The primary driver of our growth continues to be our mainstay subscription revenues.

For the full year, subscription revenues grew 13% reported and 12% in constant currency. This strong growth continues through year-end with fourth quarter subscription revenues also growing 13% reported and 11% in constant currency. Subscription revenues now represent 92% of our total revenues, up 2 percentage points from 2024. Our E365 and SMB initiatives remain solid contributors with 365 now comprising 45% of our subscription revenues, an increase from 42% in 2024.

Our smaller and less predictable revenue streams performed as we signaled during the year. Perpetual license revenues were essentially flat for both the quarter and the full year for services revenues the full year decline of 6% reported and 7% in constant currency was consistent with expectations. The fourth quarter saw a modest increase of 4% reported and 2% in constant currency. Last 12 months recurring revenues increased by 12% year-over-year and represent 93% of our total revenues, up 2 percentage points year-over-year.

Our last 12 months constant currency account retention rate remained strong and consistent at 99%. Our constant currency net revenue retention rate remains at a strong 10%. The combination of our high retention rates and new business momentum gives us confidence in the continued durability of our recurring revenue growth.

Now turning to ARR. We ended Q4 with ARR of $1.462 billion at quarter end spot rates. On a constant currency basis, our ARR growth rate was 11.5% year-over-year. The sequential quarterly growth of 4% was in line with the expectations we set reflecting our typical fourth quarter seasonality and the timing of anticipated asset analytics deals and programmatic acquisitions. For the full year M&A contribution to our ARR growth was less than 40 basis points.

Turning to profitability. Our GAAP operating income was $79 million for the fourth quarter and $363 million for the year. I've previously explained the impact on our GAAP operating results from deferred compensation plan liability revaluations and acquisition expenses. Adjusted operating income less stock-based compensation was $94 million for the quarter, with a margin of 24.1%. The strong quarterly margin expansion was in line with the OpEx seasonality we discussed in our last call.

For the full year, adjusted operating income less stock-based compensation was $430 million, up 16% with a margin of 28.6%. And this represents 110 basis points of margin improvement year-over-year, in line with our full year outlook.

Our free cash flow generation for the year was very strong, totaling $520 million, up 24% year-over-year. The fourth quarter, in particular, significantly exceeded our expectations, driven by continued strong collections and effective working capital management. While we are pleased with this result, please note that Q4 is our largest renewal quarter but the timing of collections can introduce variability across calendar years.

We maintained our disciplined and balanced approach to capital allocation. In 2025, we invested $93 million in acquisitions, while strengthening our balance sheet by paying down $135 million in bank debt and repurchasing $10 million of convertible notes. We also delivered substantial returns of capital deploying $157 million for share repurchases and $85 million for dividends.

Our balance sheet provides significant strategic flexibility. At year-end, our $1.3 billion revolver remains undrawn with access to an additional $ 500 million accordion feature. And we reduced our net debt leverage to a healthy 2.1x a 4-year low. Our current leverage range and cash generation affords capacity to fund dividends and ongoing share repurchases and up to $400 million in programmatic acquisitions annually. Subsequent to year-end, we retired our 2026 convertible notes at maturity, utilizing available cash on hand and approximately $600 million from our revolver.

Retiring this convert reduced our fully diluted share count by approximately 3%. While this repayment shifts our debt profile, we continue to actively manage our interest rate exposure. We have safeguards in place, including the low fixed coupon on our remaining convertible notes and a $200 million interest rate swap expiring in 2030.

In summary, we entered 2026 from a position of strong financial fundamentals. This provides the foundation for our outlook which builds on our long-term objectives of durable low double-digit ARR growth, continued annual margin expansion and strong free cash flow generation. For 2026, we expect our total revenues constant currency growth to be in the range of 11% to 13%. At current exchange rates, this translates to total revenues in the range of $1.685 billion to $1.75 billion.

Our mainstay subscription revenues which comprised 92% of our business are expected to grow between 11% and 13% in constant currency. In our smaller revenue streams, we expect a reacceleration in our service revenues with constant currency growth between 15% and 20%. This is attributable to increased scale of our Asset Analytics business as well as strong order book for our cohesive maximum business.

We expect perpetual license revenues to remain relatively flat.

Turning to our primary metric of business momentum. We are projecting constant currency ARR growth between 10.5% and 12.5%, reflecting momentum in our established business and that upside from our AI-powered assay analytics initiatives isn't necessarily annual recurring.

Now turning to profitability. Our long-term financial framework includes a commitment to deliver approximately 100 basis points of operating margin improvement annually. Historically, our margin percentage has had a natural hedge against currency fluctuations. However, as our business mix has evolved primarily with the growth of Seequent, which builds globally in U.S. dollars, but has significant cost in other currencies, our exposure has shifted. While a weakening U.S. dollar can now create a headwind to our reported margin percentage. It also provides greater stability to our margin in absolute dollars.

To reflect the shift towards increased stability of our absolute profit dollars we will now provide our profitability outlook as a dollar range.

Before I provide that range, I want to note one final refinement to our primary profitability metric. Going forward, our focus will be on adjusted operating income less operating stock-based compensation. This changes for consistency. Our metric has always excluded cash settled acquisition-related stay bonuses, and this refinement simply aligns the treatment for equity settled stay bonuses, which could become more significant. This removes M&A-related volatility from a key operational metric and results in an approximate 50 basis point increase in the calculated margin compared to the prior definition.

For 2026, we expect adjusted operating income less operating stock-based compensation expense to be in the range of $495 million to $510 million. This incorporates our annual improvement commitment of approximately 100 basis points in constant currency, applied to the new baseline and offset by the 50 basis point FX headwind at current rates versus 2025. We expect our effective tax rate for 2026 to be approximately 21%, consistent with 2025.

Finally, turning to free cash flow. For 2026, we are projecting a range of approximately $500 million to $570 million. The wider range directly reflects the timing variability of collections in our large fourth quarter. The midpoint of this range continues to represent very strong operational cash generation. However, the year-over-year growth is moderated as a direct result of repaying our 2026 notes, which will result in an approximate $30 million cash interest outflow compared to a negligible amount in 2025.

We expect that approximately 45% to 50% of our free cash flow will be generated during the first half of 2026. To help with your modeling, we expect our quarter-over-quarter ARR growth rate to be similar to 2025, which will cause year-over-year ARR growth rates to be relatively stable throughout the year.

Similarly, we expect our revenue seasonality to be comparable to 2025, while operating expenses we plan to invest earlier this year, which will result in spend being more weighted towards the first half and less concentrated in the fourth quarter.

I also include here on this slide additional expectations on CapEx, interest expense and cash interest, cash taxes, total and operating stock-based compensation PAUSE operating depreciation and amortization, outstanding shares and our unchanged dividend.

And with that, over to Eric to moderate Q&A.

Thanks, Werner. Before we begin, I want to point out for your modeling purposes, we have included the historical adjusted operating income less operating SBC reconciliation on Page 47 and 48 of our Q4 presentation, which you can find on our IR website. [Operator Instructions]

Our first question comes from Matt Hedberg of RBC.

Congrats on the year, and it seems like there's a lot to be optimistic for in your 2026 growth outlook I'm curious -- and I appreciate the -- all of the focus on AI at the start of the call, it's obviously topical for every investor in terms of how AI could potentially drive upside to growth. And it really does feel like you guys are well positioned there. Taking a step back, and realizing you did some smaller acquisitions to end 2025. What do you see as the most important elements that could push constant currency ARR growth closer to that higher end of the range? It feels like there's a number of opportunities. And is AI one of those?

Well, AI is contributing now in the form of asset analytics and faster growth PAUSE in asset analytics than elsewhere. The caveat is it isn't necessarily growth in ARR because we don't yet classify it as annual recurring in all cases. That's simply be because the inspections for many types of infrastructure assets are not done every year and regulatory requirement is not annual. We want to make it annual with the favorable economics of automated AI-based asset analytics. But that's a process PAUSE over time. So AI will goose our asset analytics revenues, it takes time for that to come into ARR per se.

The next question comes from Jason Celino from KeyBanc.

All right. Thank you. Sorry, it looks like I froze a little bit. So Interesting comments on kind of leverage and M&A potentially, it looks like leverage is down to more optimal level. And I think you said annually, you're open to $400 million of programmatic acquisitions. Is that $400 million consistent with what you've been open to in the past? I know you've closed much lower amounts based on your cash flow statement. So are you messaging that we could see like a pickup in tuck-ins going forward?

Well, over the past several years, while we've had higher leverage, we've been in particular to focus on asset analytics opportunities. And there we're fussy, but we've closed the ones we've cared most about. There are others of those. But now with leverage down to where it is, we are expanding our -- where we're open in M&A to beyond asset analytics potentially. And it's not that we're necessarily trying to do $400 million in annual acquisitions. But even if we did as much as $400 million, it wouldn't increase our leverage from where we are now.

Next question comes from Joe Vruwink from Robert W. Baird.

Okay. Thanks for the time today. Digital twins seem to be coming up more and more just even within the last 6 months. I know this has never been a buzzword frequently, it's a practical application and you talk today about how there's data and applications around the strategy when it comes to you. But I'm wondering if all the attention this is now [indiscernible] Digital Twins as kind of a foundation for physical AI projects. PAUSE Is that opening up pipeline? Or are you finding it's becoming easier to fund digital twin projects at the type of account that may be hesitated in the past just now that they're getting inundated with this messaging and you're certainly bringing tangible use cases that can be referenced back to them?

We're making the investments into digital twins as easy as possible because our products themselves are leveraging the digital twin technology. Meaning processes before were powered simply by files are not part by digital twins. So if you're going to do a design review right now, the technology to actually enable that the design review to be able to combine perspectives from different engineering practices is actually the digital twin.

If you're going to do a 40 model of the construction project, then the technology you're going to use to do that is also a digital twin. And then if you're going to do an inspection of an existing asset, the technology that enables to do that is also digital twin. So it's basically happening without the accounts expect being aware of because it is a technology, which is underlying our products across our portfolio there.

The next question comes from Siti Panigrahi from Mizuho.

Great I want to ask about the macro, most demand environment and infrastructure and budgets. So what have you assumed in your guidance in terms of macro? You talked about some of the trends. But the last few years, we have seen some kind of slowdown in the construction. Are you seeing any kind of changes in infrastructure owner operator budgets right now or any kind of delay projects or elongated sales cycle. And if I may, can you clarify any kind of revenue contribution from those 2 acquisitions in Q4 in '26.

So in the macro environment, we are assuming for 2026 is remarkably consistent with what we've seen in 2025, yes. So look, on our side, the only area where we've seen a slowdown over the past few years was really the sector of facilities and commercial facilities, basically buildings, but all the other sectors have been -- have grown very well. So resources, which goes beyond infrastructure, was again our fastest-growing sector in Q4. But public outs, which is the big sector we have, again, strong growth in Q4 and even industrial was solid. So that's why we're actually assuming for the rest of the year for 2026, which is a consistent demand environment.

Something as different is China, quite apart from the geopolitical aspects is slowing down. And to your second question, we don't ever break out revenue from acquisitions. We do show the because it's easy to capture the onboarded ARR from platform acquisitions, but not the revenues.

The next question comes from Kristen Owen from Oppenheimer.

Thank you for the question. I wanted to ask about your assumptions on the services revenue recovery in your 2026 guide. How much of that is being driven by some of the unique aspects of this asset analytics sort of early revenue stage versus a recovery in some of that core maximum related business? Just help us understand your drivers there.

Yes. We've seen definitely an improvement of our services business related to IBM Maximo. And other is the bulk of our services revenue. So it's really good that the investments in upgrading actually to the newest version of Maximum mass, mass 8 and 9 has definitely resumed. And that started, let's say, towards the middle of 2025 and then a carried over. And that's what we expect also for 2026.

The asset analytics acquisitions do bring us some data acquisition services but minimal. And our intention is further to minimize that.

The next question comes from Jay Vleeschhouwer from Griffin Securities.

Nicholas, I'd like to follow up on something we talked about a quarter ago on the subject of your product development and chronic releases and the SQL question is at Amsterdam at the conference, you talked about new packaging that would be forthcoming, presumably this year. Could you talk about that and whether the new packaging is it all baked into your guidance for 2026?

And definitely baked in because the new packaging was actually released in Q4. It is related to the announcements we made at our annual conference of Connect as the entry point to Bentley Infrastructure Cloud. And behind the introduction of Connect, there was a repackaging of project wise in particular, to simpler, easier to understand and consume tiers.

Now the -- we're getting a lot of positive feedback, and we got it already in Q4 about the new packaging because of its simplicity. Connect itself is getting very quickly adopted. Some of Connect comes from an older product we had called [indiscernible] 365 and about the active projects in prorate already been migrated to Connect.

So it's really taking up core of momentum, which is fantastic. But the packaging of -- and the new package of product-wise, in particular, is really resonating with accounts because now they have a much easier access to advanced capabilities such as design review or constructibility review or class detection. And it's made up for available to a much larger user base on our side. And our thinking is this is going to translate into higher consumption of those capabilities and is going to support our growth going forward.

Next question comes from Faith Brunner from William Blair.

Wanted to ask about the future of AI on the design side. You guys talked about you're playing the long game here. So maybe talk to us about how you're leveraging your network and customer base to really start designing this road map.

We definitely are not designing a road map in a very tire. We're doing it in conjunction with our accounts. Whether they are engineering services firms or owner operators, and we make sure that these accounts that we're engaging with are as representative as possible of our markets. And we are building. And I will say, 42 years of trust. Greg is using this expression of digital a quarter back. That is 42 years of us being the digital [indiscernible] of infrastructure where we've helped engineering spaces firms and owner operators successfully navigate through multiple paradigm ships from CAD to 3D, from to 40 and digital twins and now with AI.

So when we gave an update on the road map at our annual conference, we also announced a co-innovation initiative called infrastructure. And in the context of this initiative, we are talking with our accounts on how we need to evolve. The most fundamental part of our portfolio are engineering applications, both technically and commercially in order to support AI-driven workflows. We have very strong engagement, a lot of excitement, and this helps prioritize actually which are the first APIs beyond the ones we already have that we need to enable in order to support these workflows going forward.

However, as we explained in the prepared remarks, we're very clear that indeed this is going to take a while because of the perforates of the infrastructure sector for this kind of technology to be taken up. And therefore, focus is very much on adoption, more than monetization at this point in time,. But we're absolutely sure that as that technology gets adopted and as value is really created and appreciated that we'll be able to capture the fair share of that value.

The next question comes from Blair Abernethy from Rosenblatt Securities.

Nice quarter. Nice quarter. Nice way to end the year. Just on the asset analytics business, with these new acquisitions as well and more to come as you're suggesting, what's the go-to-market approach here? Is it still -- is it really just badly driven? Or are there -- I think, Greg, you and I have talked in the past about engineering partnerships or working with a large global engineering firms to try and develop a service revenue for these guys -- recurring revenue for these guys. Is that still in the works? Or just maybe an update on how you see the go to market on asset analytics.

That is the outcome intended that we will through our asset analytics platform, if you like, white label that engineering firms who will bundle it with substance of engineering services in their own IP-driven AI so that when problem is detected, the remediation is part of their own know-how and their own recurring services. We think that's something we share an objective with the engineering firms in doing that. But in the meantime, there's such a opportunity on the ground floor to get established from one asset type to another, as we mentioned, distribution poles and next transmission towers are going to be strong for us this year, and we are tending to get ourselves established there directly to start with.

Next question comes from Alexei Gogolev from JPMorgan.

Greg, are you seeing the long-term ownership to pay being impacted by the disruption dynamics? What I mean is, has it become incrementally more tricky to do a deal in light of this current uncertainty?

Deals, do you mean M&A deals?

More kind of your deals with large corporates, large customers. I'm guessing not from the ARR growth dynamics that you are posting. And it seems like the market has overreacted for a lot of stocks, including Bentley, but what are you seeing on the ground?

Well, you've heard me before, measure carefully the annual escalation baked into multiyear cap and floor extensions. And that hasn't shown any proclivity to worry about AI displacement, the engineering firms would are glad to budget spending more for technology. I think they are as poised as we are to benefit from the transformation and business model when the owner operators pay more for intellectual property through AI then for the hours of execution. And that inflection still lies ahead. But in the meantime, there's no lack of confidence as measured by our visibility into the out years of the floor and ceiling bracket that I carefully monitor.

We understand, we are, again, very early stage with respect to AI, especially when it comes to design, asset light, that's different. But when it comes to design. And I will say AI is a source of a lot of excitement and exciting conversation with our accounts. So what I can say is that it is helping us actually. For example, our commitment to data stewardship makes us quite distinct from other players out there. And for engineering services firms and also on operators to understand that if they're going to use our platform that their data remains their data always.

And we're not going to use it to train our own AI without their explicit ask actually, is really resonating, right? So it's AI is an opportunity to have a longer-term strategic conversation with accounts and for them to understand. And I appreciate that indeed, they are -- they have the right partner in Bentley Systems.

The next question comes from Daniel Jester from BMO.

Great. It's on Seequent. You highlighted fourth quarter improvement there. And if I go back, it looks like 2025, you saw pretty consistent improvement there. Maybe stepping back, if you if you look historically in the business, so we've got into an up cycle in minerals, metals and mining and the like, how has that in the past affected performance of the business? And PAUSE if we do see this continued cyclical improvement, what are the levers to drive sort of pay incremental revenue or opportunities for you.

Of course. So what we've seen is starting a couple of years ago, actually, a slowdown in big capital investments for the exploration of new mines or for major expansions to existing line. And this was related more to the overall financial environment in which mining companies have been operating with inflation rate high with interest rates high, basically making it quite expensive to raise capital. And this has improved through 2025. And that's why we're very -- I would say, cautiously optimistic about mining in 2026 and beyond.

There is clearly an upward trend here. Now what has been remarkable with our sequence business is that even during this slowdown of mining over the past few years, it continues to grow faster than the rest of the company. And that speaks a lot to the resilience of their business model. It's not so dependent on capital investments because their software is used in day-to-day operations of mines, but also of energy plans if we can -- if you want to talk about geothermal.

Now with cyclicality, you may be also referring to what's going on with the price of minerals. And that's a pretty complex topic. There are many reasons that explain why the price of coal, like any other commodity actually is going up or down, yes. But this is pretty disconnected from where the sequence software is used, which is actually on the supply, the production side, even with go, by the way, you can see that the production, the volume of production of gold continues to go up and its decorrelated with what's going on with the price of gold. And that's again because the -- regardless of what's happening with the price. Mining companies need to continue to understand the subsurface. They need to continue to derisk. They need to continue to be as efficient as possible and therefore, use Seequent software.

And an increasing portion of Seequent is for the foundation in civil and miring products, civil engineering projects because the source of -- principal source of risk is subsurface PAUSE and that has now been envied in project risk approaches, the modeling with leapfrog of the subsurface.

Next question comes from Joshua Tilton from Wolfe Research.

Can you hear me?

Yes.

Greg, I very much appreciate the master class and why Bentley is well positioned to be an AI winner. I think if I listen to kind of the questions from my peers on the call, though, what investors are trying to figure out or hoping to get a little bit more clarity on, at least from my perspective, is just out of all of the opportunities that you discussed, whether it's asset analytics or more design which of those opportunities do you see being a reality sooner? And is that a 2026 phenomenon, 2027 phenomena? And exactly how you touched on a little bit of consumption like exactly how are we going to see that monetization show up in the model. Does that make sense?

Well, asset analytics is immediate. But what's interesting is that with talent we started with API consumption. We negotiated what wound up being a 7-figure annual consumption of the engineering simulations. Based on the digital twins they captured with our iTwin capture models.

And generally, the API opportunity will massively expand the consumption and the computing for our modeling and simulation across the board. But in our experience, as we mentioned, going from CAD to BIM to digital twins, the obstacle and for reasons that I got into the institutional impediments, it's take-up that is the biggest challenge. And so we're focusing on helping users identify the APIs for which they'll create AI agents this year, we've made it as simple that they can do it now from a natural language and start to be running many more iterations and see the benefit of that. You can, in many cases, quantify how much better the design outcome will be. In future years, it will be more so reuse of existing components and constructability and so forth that we mentioned.

But it's pretty immediate to work with accounts to identify how API consumption can help them now, and we'll be working out how we monetize that with them as their digital quarter master during the year.

The last question will be from Koji Ikeda from Bank of America.

Yes, guys, can you hear me okay?

Yes. So I was wondering as we think about design software and the data that's embedded in there, how do we think about the pace of monetization potential of this data, either within Bentley itself or other vendors within the ecosystem that work with systems like Bentley to plug into your data, access that data and fervently to monetize that data. Is this an opportunity that could happen sometime in the next 12 months? Or do you think this is something that could happen potentially over the next several years?

Well, first, I will say that project-wise and Bentley Infrastructure cloud, you can see on our distribution of ARR that, that is the largest of any individual product we have in terms of ARR. So we're already monetizing the compounding of semantically addressable project data that becomes more valuable now with the opportunity for our accounts to use it themselves to -- in the ways we've described, including training their own AI models and for optimization to occur within our modeling and simulation products, but under their own AI agents, exploring that outcome space. So we are monetizing that in that respect and have been. And now over Nicholas, to talk about the future.

Yes. And to make it super clear, when we say we're monetizing it now. So this is not about using the data from one infrastructure organization and then monetizing it with all the infras organization. No, no, not at all. What we're monetizing is the ability for infra organizations to surface data from their own files in the context of product wise and Bank Infrastructure Cloud and then use it for their own purposes. Their data is there, they're a full stop. But there could be actually some additional data money, true data monetization opportunities for us going forward. I'm just thinking, for example, of season that we are like deploying across our portfolio and Cesium itself opens up an ecosystem of third-party data that can be brought into the design environment of infrastructure engineers so they can design in full context. I'm thinking, for example, of the Google 3D photorealistic [indiscernible] or some additional TERRAIN data or all the data sets that could be offered.

And I will say in the longer run, it is fairly possible that all of these engineering insights that are being created with our software in the context of opens and maintenance also becomes context for the infrastructure engineers to design, for example, extension of existing assets or do some remediation work on existing assets. And some of these insights can be created from third-party data and can be monetized as such as well. I will say, in the longer run, an exciting opportunity, monetizing potentially third-party data as well.

And it is CZM's stock and trade to become -- to have become that geospatial platform for such immersion in the world at large.

Great. That was the last question. So that concludes our call today. We thank you for your interest in time and Bentley Systems. Please feel free to reach out to Investor Relations with further questions and follow-up. And we look forward to updating you on our performance in coming quarters. Thanks.

Thank you.

Thank you.

It's really my pleasure this morning to introduce Nicholas Cumins, the CEO of Bentley Systems for our fireside chat. I think Nicholas is going to explain a little bit more about what Bentley does, that they are a frequent visitor of the Nasdaq London Conference. We're so proud to have him as a partner at Nasdaq and to have them with us here today. So good morning.

Good morning. Good morning, everyone.

How has your trip been so far?

So far, so good.

Good. Good. Hopefully, only better from here. So you've been in the CEO role for about 1.5 years now. Can you talk about any changes to the strategy or organization that you made and also the role the Bentley family plays today?

I can. Maybe a few words about Bentley first. So we like to say we are the infrastructure engineering software company. And by that, what we mean is we are a software company, and we're dedicated to infrastructure and what do we mean by infrastructure, everything that makes other things possible. So think roads, bridges, tunnels, electric grid, the water network and so on and so forth.

From a software perspective, we have software for pretty much all the engineering disciplines that are involved in infrastructure projects. So we have software for civil engineers, structural engineers, geotechnical engineers, you name it. We have software for every phase of the infrastructure life cycle for design to construction, to operations and maintenance, and we cover pretty much all the sectors of infrastructure. So that's who we are in a nutshell.

We were founded more than 40 years ago in 1984 and we went public just 5 years ago. So yes, I am the new CEO. I became CEO 1.5 years ago almost. And I'm the first non-Bentley CEO. So the company was created by the Bentley family, Bentley brothers. And I succeeded Greg Bentley, who is now our Executive Chairman, he's sitting on the Board together with 3 of his brothers, and I'm on the Board as well. So that's roughly Bentley.

So from a strategy standpoint, of course, I had been at Bentley before. I was COO, and before that, I was CPO. So I joined at the time of the IPO. So the strategy we're executing on a strategy I was already working on together with Greg, the previous CEO. So no change from that standpoint. Of course, however, a much stronger emphasis on AI, and I know we'll discuss it in a moment. So this -- it's more of an evolution, if you want, than any radical change.

And what we've been investing in, in the past few years has actually put us in a very strong position for AI going forward. So that's that. And then from an org standpoint, maybe just a couple of changes as I became CEO, we consolidated all the product teams with our technology organization. So it's all reporting into the CTO. And that's because we are in a time of great technological change, and we needed to have a very fast feedback loop between product and technology. That's one change.

And then another change is we brought all the user-facing teams together. So account management, success management, services all together under our Chief Revenue Officer. Same thing to make sure we have a very fast feedback loop across all of these functions at a time of great change.

Wow, some of the great changes. So can you talk a bit about -- you just mentioned the Chief Revenue Officer. Can you talk about demand backdrop, global funding and how you believe it will trend over the coming years?

Yes. So it's never been a better time to be in infrastructure and never been a better time to be in infrastructure engineering and never been a better time to be in infrastructure engineering software. So I'll go there step by step.

So from the biggest demand backdrop here is a lot of investments going into infrastructure around the world to support economic growth, to address climate change, to ensure energy security. And just think about some of the bills in the U.S. in the past few years, but also in Europe, for example, in Europe, to ensure as part of defense plan that we have strong dual-use infrastructure for both civil and military purposes.

So the investments going into energy transition as well. So all of that is a lot of investments going into infrastructure. There's never been so much investments going into infrastructure. Yet at the same time, there's just not enough engineers. That's why it's a great time to be in infrastructure engineering because you have a lot of work. We're not going to run out of work anytime soon. There's just not enough engineers. So you have this widening gap between demand for better, more resilient infrastructure and the engineering resources capacity that is available to do the work. And how it translates is backlogs of engineering firms that extend further out 1 year or more. You have engineering firms who are turning down jobs because they simply don't have the capacity. So they're busy.

And if you talk to CEOs of engineering firms, which obviously I do on a regular basis, they'll tell you, that's the biggest challenge. They don't have enough people. I think in Europe, there's an interesting stat, which says for the water infrastructure, about 50% of the workforce will retire in the next decade. We're now going to have 50% of the demand dropping for water, quite the contrary actually because we need to be smarter about how we go after the aquifers in Europe. We need to also tap into the water for electric generation, et cetera.

So this is not the time to have 50% of the workforce retiring. So now this is the backlog for us because what do you do when you don't have enough people, you make them more productive. And the best way to make them productive is actually software, obviously.

And we've always been in the business of making engineers more productive since we started in 1984 by telling them there's a better way to do design than a pen and paper, you can use computers to do that, computer design, going from 2D to 3D, 3D to digital twins. And now with AI, those are all interesting inflection points for the productivity of engineers.

Yes. Wow, really interesting stats. I appreciate that. Let's talk a little bit less about specifics of the business, and let's talk about results. So you delivered very consistent results over the last number of years. And can you talk about the investment thesis and your long-term financial framework.

So yes, the business is very resilient. And maybe I'll explain first like the fundamentals of the business and then we can talk about our plan going forward. So the fundamentals of business are indeed very resilient, consistent results that compound basically each year. The makeup of our business, more than 40,000 accounts now with very low revenue concentration. About half of our ARR is coming from engineering firms and another half is coming from infrastructure owner operators. That's also something to bear in mind. So it's diversified from that standpoint.

I did mention we're serving pretty much all the infrastructure sectors, but about 60% of our ARR, about 60% is in public work utilities. And this is a sector which is countercyclical as opposed to others. This is the one which is benefiting the most from these very secular investments going into infrastructure. About 92% of our revenue is recurring, about 94% of our revenue is direct. We have a very high operating leverage. We have a low financial leverage and we have a robust free cash flow. So going forward, and this is our long-term financial framework we've had since the IPO, and we see no reason to stray from it right now.

It's a low double-digit AR growth and this is composed of, let's say, mid-single digit from pricing, mid-single digit from upsell and cross-sell and about 300 basis points from new logos, which we've had actually in the past few years. So that's on ARR growth. We see a margin expansion of, on average, let's say, approximately 100 basis points every year. That's what we've been delivering for the past 5 years as well, and we expect going forward. And it's still a very robust free cash flow generation.

Yes. And you just mentioned new logos. So let's drill into that a little bit more, especially with SMB. So SMB has been a great growth driver for you for a number of years. And so how are you bringing on so many new logos a quarter and how much opportunity is still left out there in the market?

Yes. So fair point. Most of the new logos are from SMB. We have some, as we go -- continue to grow into asset operation and maintenance and we go after new owners and operators. But the vast majority is SMB. So SMB is interesting for us. About 5 years ago at the time of the IPO, we decided to be much more intentional about SMB. We used to be in the space, but more as a byproduct of serving very large engineering firms, a very large owner operators who happen to have smaller firms in their ecosystem to which they were outsourcing more. That's why we were and they were basically telling them, "You must use Bentley software and they will follow suit." That was our SMB business before the IPO. And around the time of the IPO, we said let's get much more intentional because we have the technology to do it. I didn't mention it.

One of the reasons why we're not doing so much in SMB is because we do prefer a direct engagement model. Again, 94% of our revenue is direct, right? But clearly now, there is ways of being both direct and scalable. And we've invested into our own commercial -- online commercial platform, our own sales team to do low-touch engagements with inside sales, and we're able to really scale our SMB business.

So for the past 5 years, it has contributed a lot into our ARR growth. For many years now, about 300 basis points of ARR is coming from new logos primarily SMB. And for 15 consecutive quarters, we brought more than 600 new logos to the company every quarter, more than 600 logos, all in SMB. And we see no slowdown in the foreseeable future. There's still a lot to capture there in terms of new logos, which is great.

The reaction we're getting from SMBs is it's great that you're coming to us. We like to have an alternative for whatever software they were using before. And the software, by the way, that we offer in SMB is exactly the software that we offer for bigger accounts. There is no change as well. So it's purely a go-to-market investment.

Got it. Got it. 600 a quarter, that's very, very impressive. So let's pivot and talk a little bit about data and technology. Asset analytics is a newer area that you're investing in. Can you help us understand where that business is today and your longer-term strategy?

So maybe I'll back up a little bit from this. So the -- we do have software for design, construction, operations and maintenance, operations and maintenance, or ARR in that space is less than 10% of our total ARR, yet if you really take a step back and you look at whole lifetime cost of infrastructure assets, the vast majority of that is in operations and maintenance. And that's because simply put, I mean the infrastructure assets longevity is very long. It's decades basically. So there's a lot of investments going there. And we've identified a number of inefficiencies where software can help as well. So we've been in that space for many years, if not decades, actually, but focus on asset information management, which is how do we help owners and operators maintain information about their assets. And this is a services heavy business.

And by the way, we did invest also our own system integration arm here called cohesive in order to help do this. It's a lot of integration with the ERPs of this world, the EAMs of this will, think IBM Maximo or SAP, ERPAM or Inforem, et cetera. There's a lot of system integration, which is good and it's important that we do that. But we've identified another opportunity, which was much more akin to a true software business, and that is what we call asset analytics and you were referring to.

And what is this? This is actually helping owner and operators and the people who serve them understand the exact physical conditions of an infrastructure asset in its full context. It's leveraging AI, computer vision to detect features what's going on with that asset? Is there a crack? Is there spooling? Is there vegetation growing on a transmission tower, God forbids or [ OLCM ]. And then it's leveraging our pretty vast portfolio of engineering applications to understand what all of these features mean from an engineering standpoint. So we detect that there has been more equipment installed on the tower. We can run simulation analysis to understand if the structural integrity of that tower is potentially endangered then we can trigger remediation work.

So we're quite excited about this. It's a big investment area for us from a product standpoint. And it's a big focus area for us from an acquisition standpoint. That's where we want to look for assets. We are pretty much creating a new market here together with many, many start-ups. They're not really big established players doing this, that we can go and then settle. So this is a market to be created and for us to lead.

And you mentioned briefly, you touched on the topic that we both have to talk about all day because it's so important, but AI. So there's quite a debate regarding AI and the impact on software companies. And how is Bentley positioned and how can it be a winner in the AI space?

So AI is playing a role at many levels when it comes to infrastructure and infrastructure value chain. So where we have been historically with AI is in operations and maintenance, levering, as I said, computer vision to understand what's going on with the assets.

And I will say this is relatively straightforward and relatively established, even though it's a market that we're creating, but the efficacy has been proven, et cetera. There is a big opportunity for AI in design now. And the infrastructure community overall is looking into AI as clearly a great way of solving for that widening engineering resources capacity gap on how we can make engineers more effective -- more efficient and more effective.

And so here, we play at different levels as well. So one is we are delivering our own AI capabilities to the benefits of the engineering services firms, the owners and operators when it comes to design. Some of that is next-generation applications that are developed from scratch to be truly AI native. We launched one for site engineering. We announced another one coming for the design of substation and it's quite exciting. But the majority of our users are not using this next-generation applications, of course, because we're launching them. They're using traditional applications. And so we're bringing AI capabilities now to users of these traditional applications. And we're going after the biggest time syncs we can see in the design phase.

And at this point, it's not about generating design per se. It's from the design in which infrastructure spend their time or should spend their time actually or the majority of their time generating drawings. We're in this interesting space where when it comes to design infrastructure, it's quite sophisticated, you're leveraging software 2D or 3D designs, leveraging digital twins to combine designs together, et cetera.

And then everything gets dumbed-down as we get ready for construction because we have to produce 2D sheets. We have to print them to give to the construction firms so that they can carry on the construction work. It's an interesting space. So everything gets dumbed down. This just the fact that you move from 2D, 3D design to drawings is a huge time sync for engineers. It's easily 40%, 50% of their time just creating these drawings, manually annotating them that is really the worst use of their time, right?

So think about the potential efficiency gains if all of this could be automated. This is the kind of capabilities that we're bringing to traditional applications to do that.

So now here's another interesting fact for all of you because you may have seen from many, many years ago, a stack rank of all the sectors and how digitally advanced they are and you'll see engineering and construction pretty much at the bottom, right above agriculture, which is the absolute bottom, right?

Now we've been amazed in the past few months of how much engineering firms are actually leaning on their own AI investments. Not all of them, typically the larger firms, the ones who can actually afford that, not the broader engineering community out there. But we've been very impressed. So they're truly leaning in. Why? Because they want to solve for that productivity issue. And the way we help them here is, one, tap -- help them tap into all of their past project data so they could leverage this to educate their own AI agents, so they don't have to start a new design completely from scratch. They can reuse past design from past projects.

And we can do that simply because the #1 solution that is being used out there for public works utilities when it comes to infrastructure data is our solution, our Bentley Infrastructure Cloud. And whenever you put a file into Bentley Infrastructure Cloud, we actually map that data into schemas, it can be used by AI. It's made AI ready. So that's how we help in the first place.

Then second is as they create their own AI agents, what they're looking into and it's very, very early right now. But what they're looking into is AI starting to generate some design recommendations. And they're tapping into our engineering applications, give feedback to their AI, the same way that their engineers have been using our engineering applications for decades now to be able to run simulation analysis and make sure that those recommendations that is coming from their AI agents is engineering sound.

And maybe I'll leave it there, which is -- this is still very early. It's very early, and it's going to take a while because obviously, the space in which we operate, critical infrastructure, decisions that are being made there from a design standpoint have real consequences in the real world and therefore, great care has to be taken when it comes to those decisions and the recommendations that is leading to those decisions. We need to make sure AI is really trustworthy. And that is actually what's going to dictate the pace of AI adoption for infrastructure.

So I want to open it up in a few seconds for questions. But before we do that, let's look at the other side of AI and the opportunities for your business on the large-scale data -- the data center build out. You can talk to us quickly about that, and then we'll open it up for some questions.

Huge data center buildout. As you all know, interesting to think of a data center as like a mini-city and you see data centers that have truly the size of a mini Central Park in New York or big chunk of the [ Helsinki ] Hyperion data center that Facebook or Meta is working on, is going to be even larger than Central Park of New York.

And what it means is there's a lot of infrastructure in those data centers. You need roads, to circulate around it, you have the buildings themselves. You have -- you need to be able to produce sometimes you own electricity, so you don't tap into other sources. At the very least, there needs to be electric grid in order to transmit and then distribute the electricity all the way to the data centers. You need to be able to tap into water to cool down and so on and so forth.

So we have tons of opportunities because there's so much infrastructure, tons of opportunities to -- for our software to be used and it is used in the design and the build-out of data centers. And I think our own construction modeling software is like the de facto standard now when it comes to modeling the construction of data centers. So hugely exciting.

And there's so much demand that is put on the electric grid because of data centers that it further accelerates the need for the expansion of the electric grid. So this goes way beyond the data centers now. We see that as a major driver to improve the electric grid to expand the electric grid to also push for permitting reform in the U.S., in Europe and elsewhere so that we go -- we expand that electric grid as fast as possible. So all of that is turning into a great tailwind for us.

Yes. Larger than Central Park. That's incredible. Are there any questions from the audience?

On the asset analytics piece, I understand so you're going to look at where there's a crack in the bridge or hole in the road and help people solve that. What are the firms doing now? Because what I'm getting at, is this a TCO saving for them? Or is it just nice for them to know that bridge has got to crack in or whatever?

It's a great question. So when it comes to public infrastructure, it's actually a requirement that you inspect the infrastructure asset on a regular basis. And you need to generate reports with very specific format in the U.S. at the federal level, then at the state level. You need to inspect a number of times per year, sometimes an asset, and you need to generate these reports.

The way it's done right now is all manual, yes, you send crews. So for a bridge, you have literally people hanging over ropes with pen and paper in trying to understand what's going on. There's a better way of doing this, which is what about flying a drone. And now for road, you can also send crews, but there's a better way of doing this, which is capturing LiDAR data and then processing that data and then being able to detect what's going on with those assets.

So that's what we're talking about. Now this data acquisition piece is still a services in a sense, but already more efficient potentially than again, sending people and less dangerous than having people hanging over ropes. Where we come in is once the data is being acquired from then on, everything is automated, right? So we have the data. We process the data. We understand what the data means. We detect what it is, then we understand the so what from an engineering standpoint. We generate the reports and then we trigger potential remediation work.

Any other questions?

Can you talk about your M&A strategy? Has it shifted recently more towards AI and less towards the programmatic acquisitions you used to do?

Great question because we used to do a lot of acquisitions. We've done more than 120 acquisitions.

More than double.

And we did call it programmatic because we're doing them on a such regular basis. Now we decided to focus our acquisition. It's not exclusive, but I'd say the main focus of our acquisition investigation is around asset analytics, looking into companies that are like start-ups that are already breaking ground there because it does take time to train the AI model on specific asset classes and specific features that need to be detected, et cetera. So we're quite active there. And that's why you should expect us to do acquisitions. But it's not exclusive, right?

So we'll continue to see if there are some white spaces we need to fill in with our core engineering applications, we'll do so, for example. But the main focus is on asset analytics. Now because it's a new market, and we're talking about start-ups, we're quite rigorous about the kind of assets we will be looking to acquiring, and that's why you don't see them coming up so often.

[indiscernible].

No. The most important for us is the technology indeed, yes. And they're small. So they will not be profitable, but will absorb the dilution impact in our margin expansion.

Great. Well, Nicholas, this has been a fantastic chat. We have 2 more minutes. So before we let you leave the stage, we'd just love to hear your outlook for 2026 and what investors should be thinking about in terms of the environment that you're planning for and your ability in '26 to achieve your long-term goals. So if you could leave us with that.

Great. Yes. I'll go back to my very first comments. It's -- there's never been a better time to be in infrastructure, infrastructure engineering, infrastructure engineering software. As we commented in our last earnings call, the demand environment remains very strong. We don't see any change to that demand environment going into 2026. So we see no reason to deviate from our long-term financial framework of low double-digit ARR growth, margin expansion of approximately 100 basis points and robust free cash flow generation.

Perfect. Well, thank you, everyone, very much. Thank you so much, Nicholas.

Thank you.

All right. Making our way through the morning. We've got a really great stretch of companies here, and we're continuing the trend here with Bentley. We've got Greg Bentley, the Executive Chair. So used to calling you CEO, but the Founder -- Co-Founder and Executive Chair, Greg Bentley.

You've been the Executive Chair, I think it's a little over a year now. Is that about right?

Yes.

But you're still actively involved. We were just saying you're still actively involved in the company and capital allocation decisions and broader strategy.

Could you just give us a sense for, I guess, what are you doing these days at Bentley? And what are you excited about now as you look to the future of the company?

Well, Matt, by the way, thanks to each of you for your interest and attention here today. As Executive Chairman and being a finance guy by background, I expect to always be involved in capital allocation. But the fact is that I enjoy meeting investors. And for the time being, I'm hanging on to a role in Investor Relations as well. Of course, our -- Bentley's -- are the majority of our Board of Directors and take an active interest, I'm the oldest of them.

And I have had the question, should we expect that there's a transition in ownership of the company from the Bentley family interests to other investors? And I say, well, no, I don't necessarily expect that myself. Since the company went public 5 years ago, on average, our family's holdings when we checked have gone down by about 1% of ownership per year, but still are the majority, and we're very enthusiastic about our opportunities and are very much aligned, therefore, our company's ownership with the -- with those of our long-term investors.

And of course, what there is to be excited about is that we have a new generation of management, literally, our CEO and other C-level folks. We worked hard to recruit and develop people who are in their 40s [indiscernible] and look forward to their generation of stewardship of the company in time.

Yes. Well, you took my second question. But -- so I asked you this yesterday and you said, so the answer on family ownership is kind of similar to what it has been at this point.

I don't know otherwise.

Yes. Yes. It certainly feels like you guys are -- you have built and continue to have long-duration growth and margin expansion opportunities. And so that certainly has been consistent there. And I guess to that point, you guys have delivered very consistent results this year.

We always say it's sort of wash, rinse, repeat with Bentley. It's the consistency since your IPO has been remarkable. When you sit back and reflect on -- and you've got such a diverse global business with multiple brands and multiple theaters, can you talk to us about sort of the demand environment right now, just at a very, very high level from your perspective, and then we can drill down on that further?

Well, what's consistent across the products in the world is that we're the quarter masters, if you like, providing the digital tools for civil engineers, structural engineers, geotechnical engineers and those who work with them and the end market demand for their work has never been higher.

And it's literally better each year in terms of their backlogs and there's a resource capacity shortage of engineers in the U.S. The engineering firms on average, have 1 out of every 9 positions unfilled. So the consistency in our business corresponds to the consistency of requirements for infrastructure engineering in the world.

And developments on the margin are only emphasizing that. If we just talk about this year, you have the impulse for each country in the world to be more self-sufficient in terms of critical resources and defense, requiring redundant infrastructure expenditures if it came to that. And as we sit here, the Climate Summit in Brazil has appropriately turned its focus realistically to adaptation as being needed for inevitable climate change, and that's the work of civil and structural and geotechnical engineers.

And in general, we had a country in the Western world this year where the lights went out for a couple of years, just a reminder that quality of life depends on the quality of infrastructure resilience. And as it so much ages, there's more work to be done to keep it fit for purpose and so forth.

I know we'll talk about AI being able to help with that. But if we talk about maybe tying in the subject of consistency in our end market with our operating results at Bentley Systems, there's some -- there's a way in which it coincides our enterprise accounts on our E365 pure consumption program, which is the majority of our ARR now, pay us purely for consumption.

We don't have revenue and ARR except as they use our products by the day. Now of course, they pay us once per year, so there's no change to our cash flows, but we -- there's nothing we account for on a multiyear basis and so forth. But these accounts are relying on going digital to make up the resource constraints in engineering capacity and have tended to ask for ceilings on their consumption charges under E365.

And we said, yes, we're agreeable to that, provided you will grant us a floor that began a few years ago and has become the norm. It's not a condition of our contract, but it happens to be the basis of the negotiation now each year in general. And the accounts have said, well, actually, we have other vendors for whom we lock in a multiyear contract. Now that's not good because it doesn't matter what we consume. It may be shelfware for us. So we'd much rather have consumption. But could we look out and bound our consumption with floors and ceilings over 2026 and 2027 and 2028 perhaps. And we have said this gets to the nature of the ownership and preference of our company.

Well, we depend on and are committed to double-digit ARR growth. But if the account is willing to escalate the floor and ceiling each year by an amount in double digits, we don't mind having more visibility into the future that, that would result in. And so that now is the majority of our ARR dollars from E365, which is the majority of our ARR, that there are multiyear escalated floors and ceilings that get negotiated each quarter in the renewals for those accounts.

And that gives us a glimpse into the expectations. It's a new glimpse each quarter because it's a new set of negotiations, and it's all over the world, and it's all disciplines and end markets for us. But it tends to average about 10% that engineering firms say, yes, if you'll bound us a 10% more spend more each year, we're happy with that, and we'll grant you the 10% escalation in the ceiling.

So that is -- we could be a firm that would say, "Oh no, no, we may be giving up some upside. But we're in our 44th year. We're happy with double-digit ARR growth. And this gives us kind of a flywheel of predictability, which has been the case this year. Now I say with that flywheel on the other hand, that positions us well to be willing and able to take on some more volatile new business opportunities, such as the asset analytics business where we charge per asset rather than per user or per usage.

And that is much lumpier because we're competing for big procurements. We also have in our ARR growth, the coming and going of ARR with -- that we onboard with acquisitions, and those have become much less predictable as we strategically preferred to focus on the opportunity for asset analytics acquisitions. So those have become episodic and added volatility to the total ARR growth, but the ARR growth from the from the mainstream of our business has become more predictive.

Yes. And I guess, do you envision a point in the future where -- I mean it was somewhat depend on E365 penetration, but what percentage of your business could be consumption in the future? Could it be -- maybe it's not 100%, but is it 90%?

Well, for all the enterprise accounts are about 2/3 of our business. And so far, we're about 2/3 in the process of upgrading them onto the E365 consumption subscription. Previously, they owned perpetual licenses and paid us for subscription on that. And we started with the largest accounts, and we're working our way down. So we have several years at the current rate remaining to get all of the enterprise accounts.

We count an enterprise account as those who spend $100,000 a year or more with us. Those are accounts engineering organizations of the scale to have central tool smiths who are choosing the tools. And with 365, we embed our own people into the company, into the E365 account for the year to help them instill new digital workflows and expand our tools and competitive displacements and so forth, but they have to have a central group that we can work with, and that's about at $100,000.

So we have years ahead of getting around because there's a capacity constraint on our part. We have to provide the staffing to dedicate these experts of ours, civil and structural and geotechnical engineers into the each E365 account has named Bentley success people on their team, and we have to upskill people to provide that role. And so we have years left before we'll get all of the enterprise accounts on consumption.

Now of course, the mix of consumption will start to change and include new types of consumption. But that will, I think, be the bulk of our existing business where we charge per user and usage and then we have this new opportunity asset analytics where we charge per asset.

When we think about the durability of that 10% growth, the question that we often get is what are some things on the margin that could -- IIJA was certainly part of a growth element several years ago. But when you think about some of the big moving pieces, whether it's federal funding, permitting reform, asset analytics. What are some of these bigger elements that could be some of that incremental growth beyond that sort of that floor and ceiling increase on an annual basis?

Well, and recall that the 10% I'm talking about there is for the enterprise account, mainstay of our business, but we have managed an additional 3% of ARR growth from SMB, the new prospects for us because we didn't -- we're 93% direct sales and SMB is a new focus for us since IPO, and that's providing the other 3% or so.

But across the world, governments have greater commitments to infrastructure spending, public works and utilities being the largest end market sector, for us, followed by resources. And in each of those, the requirements for the reasons of the sort we've been talking about are greater than ever.

So in the countries of Europe and the U.K. and Germany and at the EU level, new spending commitments and programs the Middle East always strong, but stronger than ever. And Asia, Australia has had a slowdown in transport spending. But now with the Brisbane Olympics in 2032, a very strong new program starting. India is committed to -- they say, they're developed economy 2047 program, which is mainly a program of infrastructure spending.

And I might say that speaking of developing economies, we showed some market research that as a country would advance and grow a quintile in GDP per person the infrastructure engineers double what they spend for the year and so forth. So that's a source of future growth for us as well. But finally, this notion of critical resource constraints, and I suppose you could say, by the way, electrical power is a critical resource. And it isn't, I think, even so much that the demand for data centers has caused this.

Our grids are overstressed and overaged and we haven't added new capacity. So finally being able to get past the permitting obstacles to -- for both grid expansion, new capacity corridors for transmission and new mining exploration.

Those are 2 major opportunities of ours. Our sequenced subsurface modeling business, which is now more than 20% of our company, the power line systems acquisition, the world's tools for physical design of the transmission and distribution grid, those already are our fastest-growing parts of the company and are poised to grow faster. In the case of Power Line Systems, it already has doubled since we acquired it 3 years ago.

I said at the time, there aren't many aspects of our business that could grow by an integer multiple. This is one. The integer so far has been too, but with the opportunity for permitting reform. Of course, that won't happen immediately because even though in the U.S., we may have permitting reform legislation, that won't mean that the permits are automatically -- It takes time. But that looks like another opportunity for the future.

I have to ask, we've been talking about IIJA for years and years, and is set to expire next year. What are the prospects from your perspective on the next iteration of that? What is sort of the kind of the scuttlebutt in the industry on it?

Well, certainly, the engineering firms monitor it very closely. A technical change will be that the IIJA was the first federal infrastructure spending, which went beyond surface transportation to add funding for broadband, grid and water because those traditionally have been funded privately or at the state and local level in the United States. That aspect is unlikely to continue under this administration.

However, broadband, grid and water are abundantly funded at the private level in the United States. So the successor to IIJA will be a regular every 5 years, surface transportation funding bill funded primarily by the gas tax. And what's, I think, most gratifying about this is that I think in the U.S., after IIJA, we no longer say the word stimulus after infrastructure.

We recognize it's essential for quality of life, and there has been a good returns on the investment. We have some infrastructure we can actually be proud of instead of a ashed of in the world. And certainly, electrical capacity is a case in point of something for which there is no substitute. But in the replacement surface transportation bill, the discussion is of baseline levels that will start at the IIJA level for surface transportation and go up from there.

So I think we've learned that infrastructure is a great investment. If I go back to the other countries in the world that I talked about -- when we talk about Europe, especially, there could be a concern maybe in this country, too. Governments are going to run out of the money. They're not going to run out of their intention and responsibility to provide infrastructure, but what about the fiscal capacity.

And of course, the answer to that is there's more than enough private investment that would like to be invested in infrastructure and it's bound to happen and is the -- when it does happen in the future because they've got the impediment now is governments not allowing private investment in public infrastructure. When it does happen in the P3s, for instance, in Canada and Australia, it's always a design build to operate, maintain, concession and the consortia that bid on that are going to win the project and make money because they have a digital twin approach in mind to utilize the engineering designs throughout outpatients [indiscernible] as a competitive advantage. So that will actually accelerate going digital infrastructure engineers.

Well that's a good segue way into AI. I mean you guys have been focused on AI for years, and digital twin has been a big part of your strategy. How should investors think about how Bentley is leveraging or embedding AI into their platform to enable customers to see even more benefit in advanced models in the future?

Well, an efficient way to look at that would be to view on our website, our CEO, Nicholas Cumins, keynote last month at our [ Year End ] Infrastructure Conference, where he spoke about that for an hour and gave some demonstrations. But you're right that AI isn't new for us. The digital twins tend to start with reality modeling from overlapping video in a drone.

And when you process that into an engineering ready reality mesh, over the past decade in our products, you use computer vision, AI to recognize and classify the digital components. Our engineers have used our generative components for scripting that can be automated to create and explore a whole universe of design alternatives and in something like water systems operations using neural networks to locate and predict leaks have all been uses of AI in the past. But now we have AI native products.

And as an example of the additional value they can create our open site plus for the simplest of civil engineering problems, a site design, here we optimize the cost of grading and drainage by running tens of thousands of alternatives. And in general, average about $10,000 cheaper per acre than would be the best design than an engineer would be able to do themselves, interactively specifying and alternative even costing it out.

So if you let AI help with the optimizing, you run the software a zillion more times, but you end up with an improvement in quality of the result. So that's a particular unusual case where you can measure the value in most of infrastructure that's more complicated. There's more factors, you can't quite do that. The value produced by AI is comparable, but you can't quite measure it in the same way.

The greatest advantage that AI is going to bring an infrastructure engineering is to reuse existing design -- designs informed by how well they've performed has tracked through a digital twin doing operations and maintenance. And that's not something that existing, that human engineers are very willing to do the work to do. but AI is great at it. And our new Bentley Connect offering foundation for our Bentley infrastructure cloud of project-wise, asset-wise and Syncro will provide an entry-level opportunity for users to federate and search and query their existing data.

Because it's their own data that they can use to train their own models or refine the training of models that we provide, but in particular, so that the economics of each product are improved by starting on what's already been designed and proven and the quality will be better and the risk will be low.

Is there an element of a network effect because you implied that it's their own models that they're being trained on. But is there a community effect that I don't -- if it's some randomized data across customers that can be sort of a better provide uplift to the entire platform? Or is it -- is there no element to see kind of a network effect?

Well, we are very clear that our users' data is their data. And intellectual property concerns are the principal caution that engineering executives have about AI. We don't have any of our users' data, it's their data. We will provide the tools for them to train AI models on that data because it's all intently infrastructure cloud for their own competitive advantage.

And to the extent the data that an owner, the Department of Transportation or airport has rights to that data, they will be able to use that to train a model specific to those assets. But it's only if they would contribute it to be used in training that we would do. And we even have a particular pedigree that we can track what data we have used.

But it's our attitude that by virtue of Bentley Infrastructure and project-wise, which has captured the engineering projects for most infrastructure engineers for their lifetime that, that can be and should be an advantage to them in their own development of their asset analytics opportunities themselves. For future projects, to reuse that data and to provide with asset analytics, the operations and maintenance logic and optimization for the assets that they've designed.

So from a monetization perspective because the question is like, it's great that you guys are focused on adding more AI capabilities, ultimately, it should drive higher consumption over time. Is that sort of the premise on providing some of these extra features?

Well, it won't drive more consumption per user per day, will it because the user will get more done in a given day. So the usage it will drive is what I call API usage. So AI used through an API where our enterprise accounts are investing in AI. They aren't investing in AI to replace their engineering calculation tools. They wouldn't take the risk of doing that. They don't need to do that.

They'll surround applications like ours with agents that help a user and engineer get more done without having to sit there in front of the screen while it occurs. And there will -- that is an example, as in the example I provided before, will greatly increase the total consumption. We'll end up monetizing, not only attended consumption as we do now, but also the API consumption and we're now working with our users for how that would be fair. In the meantime, we're encouraging as much of that API consumption as they will do because we will be monetizing it in the future.

I have about 7 more questions, and we have 4 minutes to go here. So we'll try to do something quickly. I mean you mentioned asset analytics earlier. It really does seem like a driver for '26 and beyond. How should we think about asset analytics contributing to the model? It's a little like you said, it's lite different pricing mechanism, but...

Well, where this came about, digital twin makes sense. All of the engineering logic that the owner pays for only gets used traditionally once during the project delivery, the original -- after construction, it's abandoned, whereas the structural and the civil and mechanical logic should be used to optimize the maintenance of the asset over time and make sure it's still safe and only do the necessary maintenance, spending less to do only the necessary maintenance is the driver that makes digital twins cost nothing or negative and is an opportunity for everyone.

But if you had to start with a new project for a digital twin, that can be a long sales cycle. With asset analytics, you can have an instant on digital twin with the result of a drone flight for instance, for a cell tower or in the case of Roadway miles, our blinks uses AI on Dashcam data that's available anyway every day from the fleets that are transiting the road to recognize what needs maintenance.

So you can turn on a digital twin immediately with asset analytics. And we use the term because we charge per asset per mile or per asset per year, and it's -- they will end up being an asset analytics solution for every infrastructure asset over time. We would like to have our platform be the foundation for the cloud services to do that, and to have the engineering firms adding their own trained models, their own engineers who won't need to spend the time. They've been spending instead can apply themselves to optimizing the maintenance and to add their own bundled services to our application layer.

So our capital allocation priority for the past couple of years has been to do acquisitions that add special asset-specific logic to our platform. And we haven't done one yet this year, but we expect to conclude the year having succeeded in our aspiration for really owning the space for asset analytics.

And then maybe just from a broader capital allocation perspective, you -- it's been a while since you've done seen -- what -- from your -- I mean, and you're actively involved in capital allocation decisions right now, especially from an M&A perspective. What are the prospects of a larger deal similar to those 2 priors?

Well, I say that -- we Bentley Systems would like to continue and expect to continue to be a consolidator. And this example of analytics is a case in point. There are many worthwhile start-ups that have focused in particular asset types. But to have a whole broad platform for the digital twin is beyond what they should be working on. And putting those together is our capital allocation priority for the moment. And we have many good prospects underway.

And maybe just to close, you've been involved with Bentley for over 40 years. When you think of big moon shot or, boy, 10 years from now, I would be -- that would be an audacious goal for Bentley. What were some of the big dream to dream moonshots that you think about for Bentley?

The opportunity is in asset operations and maintenance to use the engineering knowledge and intelligence to optimize and spend only what's necessary on maintenance to monitor with continuous surveying continued safety and to improve the performance and lifetime and fitness for purpose. That fortunately is the vision of our new management generation. And it's orthogonal to -- and an opportunity in addition to our existing per user business, but this per asset business is our destiny foreseeable.

Yes. Yes. The S curves of growth seem -- it's got some duration to it...

And we're going to reboot the original curve for Bentley Systems in this asset analytics opportunity, I think.

And doing it in a very Profitable way from a margin expansion.

I'd go so far as to say it's what we should do because our business is otherwise so predictable. The end markets are so stable and so forth, we should be taking on some risk to go for this. Our partner, Google is providing a lot of the data needed and it's very exciting.

And so -- and with all of that said, you still think from a predictability standpoint, should think about kind of 100 basis points of margin expansion despite some of these investments?

I say about margin expansion. When you have it in grain that we get more efficient by 100 basis points per year and everyone expects that. And we accomplished most of it through our scale leverage of being direct sales, if you see what I mean. If you had indirect sales, you spend the same amount on go-to-market every year as a percentage. With direct sales, our enterprises spent 10% more, and our cost of going to market is 3% or 4% in our annual raises that falls into the bottom line.

When you have that ingrained, you don't try to renegotiate it every year because there's always new opportunity, and we're better simply to stick to this. We have managed to compound our free cash flow per share while public for the 5 years at 14% plus. And we expect to be able to increase that next year with reducing our share count a bit as we retire our convertible debt.

And when you do that, you compound the basis for valuation doubling every 5 years, as we have done. And probably, as we have done for 40 years, and I hope to do so. And it's our plan and my responsibility for the future. I make sure I have.

Well, from all of us at RBC, thanks, again, Greg, and thanks, Eric, for supporting us.

Cool.

Good morning, and thank you for joining Bentley Systems Q3 2025 results. I'm Eric Boyer, Bentley's Investor Relations Officer.

On the webcast today, we have Bentley Systems' Executive Chair, Greg Bentley; Chief Executive Officer, Nicholas Cumins; and Chief Financial Officer, Werner Andre.

This webcast includes forward-looking statements made as of November 5, 2025, regarding the future results of operations and financial position, business strategy and plans and objectives for future operations of Bentley Systems Inc.

All such statements made in or contained during this webcast other than statements of historical fact are forward-looking statements. This webcast will be available for replay on Bentley Systems' Investor Relations website at investors.bentley.com on November 5, 2025. After a presentation, we will conclude with Q&A.

And with that, let me introduce the Executive Chair of Bentley Systems, Greg Bentley.

Good morning, as the case may be, and thanks for your interest and attention. I'm pleased to say that all quantitative metrics for Q3 are quite in accord with our expected progress and outlook range for the year.

But this quarter, Nicholas will highlight the significant product announcements and developments presented and observed at our year and infrastructure 2025 Conference last month. which I think also merit your firsthand review at the link here.

Now I always look forward to discovering through submissions for the annual Going Digital Awards the unanticipated ways by which our users are ever creatively applying software and cloud services. This year, I was pleasantly surprised by the plurality of those citing contributions from AI. So upon observing this AI forward propensity at the level of projects and users, I reviewed with interest this year's AEC advisers survey of engineering firms participating in their annual CEO conference.

You may recall that I previously reviewed 2 earlier such conferences, where Bentley Systems helped with gauging progress and appetites in Going Digital. The surveyed firms together perform most of the contracted infrastructure engineering outside Asia with the distribution of their revenues by sector weighted like ours in favor of public works/utilities and resources.

And within general building corresponding to what we classify as the commercial facility sector, the survey highlights a dramatic and interesting transition. AEC firms are now literally engineering the infrastructure for AI as spending for construction of data centers, such as the project by digital construction leader, DPR, which served as the example, throughout our year on infrastructure keynote presentations, ramps to soon overtake spending on office spaces.

AEC advisers shows that digital investment as an internal priority is also succeeding for engineering firms. For the last 5 years, they, in aggregate, have achieved continually increased profit margins at the same time as also higher growth in organic net revenues, the latter perhaps limited by capacity constraints as separately reported backlogs reached record levels.

Underscoring market robustness, this organic revenue growth is still increasing, including through 2025 estimates and net of both annual U.S. inflation in red, and in blue U.S. GDP growth. AEC advisers concludes that this growth in aggregate profit margin must be attributable to improvements in direct labor productivity as the total revenue percentage of other costs to support functions has risen continuously by almost 20%. This is despite real estate costs having declined since prepandemic by 25%, presumably owing to virtual and hybrid working enabled by our ProjectWise and other cloud services technologies.

And most significantly for us, these firms overall technology spending as a percentage of revenue will have increased by 40% over the 6 years through 2025. Combined with their organic revenue growth, their technology spending in dollar terms increased from 2019 through 2024 at a compounded annual growth rate of 13%, tolerably coinciding with the growth rate of Bentley Systems revenues, as I have reviewed in recent quarters, over our 5 years as a public company.

I believe that we have thus effectively enabled AEC firms to keep up with accelerating demand despite now chronic engineering resource constraints by constantly improving their labor productivity through Going Digital.

To understand changes now underway in the makeup and magnitude of AEC technology spending. This year, we again helped AEC advisers with a supplemental AI survey yielding sufficiently representative responses. In the interest of validating the prevalence of the commendable self-help AI initiatives that relatively surprised us within this year's Going Digital Award submissions. We focused survey questions on AI that these AEC firms are already implementing, not just testing to support their businesses.

Excluding for this immediate purpose, more widespread AI implementations for generic business purposes such as finance, HR and legal, about 1/4 of responses report AI already being implemented around the periphery of applications such as ours to support the infrastructure engineering-oriented functions of design automation, construction planning or monitoring and/or asset performance and maintenance.

Asked in what respects competitiveness would be advanced through faster AI adoption. These firms expect superior project delivery and quality, operational efficiency, and clients' experience and satisfaction, but they have the greatest regard for AI's potential enablement of innovation and new services.

To get to these benefits, the median reported level of AI implementation spending today, ranging from 6 to 53 is 19 basis points of gross revenue. That's on the order of 5% of the overall technology spending rate we just reviewed and as a frame of reference, this already somewhat exceeds what such firms on average are spending on all of Bentley Systems offerings.

Most significantly for us, these firms anticipate increasing their annual AI implementation spending over the next 3 years to a median ranging from 35 to 164 of 71 basis points. A multiple of almost 4 from today. If all other technology spending would just continue to grow at the same rate as over the last 5 years, this projected AI increment would cause total technology spending as a percentage of revenue to grow about 50% faster than at present.

But we know the resulting AI impact will be such that rather than so extrapolate, we need to factor in the probable AI accelerated changes in infrastructure engineering business models as innovation and new services are enabled. This was the subject of dialogue with the diversity of thoughtful marketplace participants, including public and private sector infrastructure owners, as we helped lead a separate survey and convened an in-person discussion in September in London that culminated in this white paper, the impact of artificial intelligence on the built environment.

The majority of the 140 senior opinion leaders survey expect the impact of AI on current business models, either to augur a major disruption and so are already taking steps to adapt or to impact to a significant extent. Interpreting the qualitative feedback as well, the knowledgeable white paper authors venture that AI will finally catalyze the long-awaited tipping point and engineering business model mix from hours-related revenue towards value-priced data-enabled services and performance/outcome contracting.

To be sure, the emerging opportunities accordingly anticipated around automation, analytics and digital twins bode well for Bentley Systems forward-looking initiatives. But to the extent that our accounts would become incented and through AI increasingly able to more so minimize their currently generally billable engineering hours and days because they would instead be variously fixed and value and outcome pricing, what could be anticipated about the consumption of software and cloud services underlying our own business model.

I could describe what we currently measure as consumption attended by a user, and that's charged within E365 per day for our open applications and per quarter for ProjectWise and most term licenses. As our AI native plus generation of applications progressively roll out, the commercial norm for our attended consumption charging is likely to become a hybrid combination of these factors and of surcharges based on computing intensity.

With our AI accelerating the pace of engineering productivity growth, attended consumption should generate commensurately higher value and hybrid monetization per relatively slowing time and/or frequency of attended usage.

At year-end infrastructure, Nicholas previewed the commercialization of an already evident source of incremental consumption with our application engines accessed through APIs to provide essential engineering context for simulations and analytics programmatically invoked by our accounts and users, AI agents.

By virtue of our ingrained platform orientation, we are very enthusiastic about working with our enterprise accounts to prioritize development of many further such APIs and to arrive at reasonable monetization for the burgeoning value that API consumption will generate. Among potential AI-enabled business model innovations, the sited AI survey show me that engineering firms and owners share our asset analytics aspiration. For digital twins created and curated through AI to become the foundation for infrastructure inspections, operations and maintenance.

Bentley Systems is investing resolutely to lead this charge internally and through our ongoing prioritization of capital allocation for pertinent strategic acquisitions. With critical mass for escape velocity gathering I believe the resulting asset consumption will become for us another mainstay of subscription revenue growth, not only within owner operators, but also as their digital integrators with co-innovating engineering firms.

My expectation for the confluence of our maturing incumbent consumption model and these new and incipient consumption streams is influenced by the way that these surveys and how our enterprise subscription renewals show that infrastructure engineering executives are assessing against the backdrop of their engineering resource constraints.

Their current combination of record margins, organic real GDP plus growth and backlog and there -- auspicious opportunities in the infrastructure AI transform future. In the short and medium term, the prevailing sustainment of our E365 renewals, including from multiple out years, at negotiated annual floor and ceiling escalations consistently averaging about 10% reflects shared confidence of enterprise accounts in our Bentley Systems and the continued healthy overall ingredient of a changing mix evolving to everyone's benefit of attended API and asset consumption.

And now to review, as usual, are our robust markets and execution, including also notably strong SMB and new logo growth and to highlight our year-end infrastructure announcements and feedback, over to Nicholas. Thank you.

Thank you, Greg. A few weeks ago, infrastructure leaders from around the world gathered in Amsterdam for annual year Infrastructure Conference to showcase excellence in infrastructure delivery and performance through digital innovation. Amsterdam celebrating its 750th anniversary is a city built on land reclaimed from the sea through generations of engineering ingenuity.

It was a filing stage for YII Going Digital Award, that same spirit of innovation took center stage YII was also an opportunity to share progress on last year's key announcements, such as integrating Cesium and Google Geo data across our portfolio. But today, I will focus my remarks on infrastructure the theme introduced by Greg.

The backdrop remains the same, whether to address climate concerns, ensure energy supply or more broadly support economic and population growth. Our world -- unprecedented demand for better, more resilient infrastructure, yet lacks the engineering capacity to deliver it. We must make existing engineers more productive by empowering them with better tools, smarter workflows and more connected data.

At YII, the Going Digital Awards finalist once again showcased how Bentley Software helped them achieve meaningful productivity gains often in the range of 15% to 25% or more. These gains were impressive, most advanced projects and don't reflect the industry as a whole. Scaling them across all projects will help narrow the gap between global demand and current capacity, but closing it requires a step change in productivity.

That step change is just beginning to take shape, and it's AI. The AEC advisor survey referenced by Greg those large engineering firms, making substantial investments in AI for design automation. For those building their own agents, Bentley can support them in several critical ways. First, we help them tap into past project designs. Every infrastructure asset is unique, but new designs shouldn't start from a blank screen.

Historically, design data has been trapped in different for formats and proprietary systems. Pending infrastructure cloud, part by iTwin data is ingested from a wide range of power formats and map to our infrastructure schemes so that it can be queried, analyzed and reused, including by AI.

In this context, we announced Connect, a new financial layer to Bentley Infrastructure Cloud. Connect delivers a connected data environment for project and asset information, improving collaboration across the entire infrastructure life cycle. From there, product-wise, for design and construction workflows and asset-wise for operations and maintenance. Connect will be generally available in December.

Next, we help firms to create their own AI agents by providing engineering context ensuring their AI recommendations are grounded in sound engineering logic and physical principles. Hyundai Engineering, a Going Digital Award winner in 2023, demonstrated this by using our stat simulation application to -- integrity of AI-generated designs.

This year, I highlighted 4 similar examples in my keynote. All drawn from an even larger number of Going Digital Award submissions that illustrated how Bentley applications provide engineering context to AI. Infrastructure Engineering is accretive profession but one where precision is nonnegotiable, and consequences are real. The same way that infrastructure organizations have trusted our broad and patients to empower their individual engineers. They are turning to our applications to provide the same precision to their AI agents.

Now as our applications were not designed to interact with AI agents, we also announced the infrastructure AI co-innovation initiative, inviting our users to partner with us to explore how our applications need to evolve, both technically and commercially, as Greg mentioned, to better support these AI use cases. At [ YII ], we also highlighted the AI capabilities we are delivering to the broader engineering community, starting with our next-generation applications powered by AI.

Open sites announced that last year's YII for site engineering is now in limited availability. We also introduced 2 additional next-generation applications in early access this quarter. Substation plus for collaborative substation design and Cinco Plus for 40 construction modeling with AI-driven insights -- patients feature Bentley copilots, our AI assistant purpose built for infrastructure engineering.

We are also enhancing existing application with AI, bringing Bentley copilots and AI-powered drawings production to open roads and open rail. And we unveiled new search capabilities in Bentley Infrastructure Cloud, powered by AI as demonstrated on stage with ProjectWise. One last point. We talked about how engineering firms are leveraging our software to ensure that the recommendations from their AI agents are trustworthy. A related topic is trust from the engineering firms in the data that we use to train AI capabilities.

The AEC advisor survey shows security and data piracy has the top concern of engineering firms with respect to AI. And this is across all firm sizes. At YII, we reaffirmed our stewardship first outlined 2 years ago. We expect for intellectual property is foundational to Bentley's approach. Users control their data always, they decide if and how it is used for AI training.

To uphold this principle, we implemented strict governance. Only data explicitly licensed or explicitly contributed by accounts for the benefit of the broader Bentley community. Users can also fine-tune Bentley AI models with their own data for their exclusive use.

And to ensure transparency, we introduced the data agreement registry, an auditing system that shows exactly how data was used to train Bentley AI models. When others are vague on these critical topics, we lead with clarity. Overall, we were pleased with this year's year in infrastructure, receiving great feedback about our comprehensive and principal approach to infrastructure. And I encourage you to check out our sessions and going to total award winners at yii.bentley.com.

Moving on to our results for the quarter. We delivered a solid quarter in line with our expectations. Our year-to-date results position us well to finish within our outlook ranges for the full year. Low double-digit AR growth, approximately 100 basis points of margin expansion and robust free cash flow, consistent with our long-term financial framework. Q3 ARR increased 10.5% year-over-year or 1% when excluding the impact of China. Growth was underpinned by a net revenue retention rate of 109%.

E365 performance remained solid, and we added 300 basis points of ARR growth from new logos again. primarily within the SMB segment. For the 15th consecutive quarter, we added at least 600 new SME logos through our online store with retention in this segment remaining high.

Turning to our -- sector. Resources was once again our fastest-growing sector in the quarter. We continue to see soft signals of improvement in mining exploration. Public works and utilities delivered another solid quarter, consistent with first half performance and driven by sustained global infrastructure investment. Power Line Systems remain a standout performer benefiting from global demand for grid resilience and increased power generation. Growth in industrial sector remained modest, to facilities was flat.

Looking at our geographies at a high level. Asia Pacific had a strong quarter, followed by the Americas and EMEA. Growth in Americas was solid, led by North America. In the U.S., our accounts continue to benefit from a favorable macro backdrop despite ongoing uncertainty, though less so from tariffs and policy shifts and the recent federal shutdown. To date, we have seen minimal disruption from the shutdown.

Looking ahead, there are a full-scale promoting reform for energy infrastructure and critical minerals in the U.S. will happen in the coming quarters. Both our Power Line Systems and Seequent businesses are very well positioned to benefit from these governments. In EMEA, the Middle East continued to lead the region with another very strong quarter, followed by Europe and the U.K. Long-term opportunities are supported by robust investment in transport, water and energy, particularly in areas such as dual use in passenger expansion and nuclear.

There's also movement in Europe on permitting reform. The European Commission published guidance to have member states accelerate permitting and deployment of renewable energy and grid infrastructure as part of its broader effort to lower energy costs and strengthen supply security.

In Asia Pacific, overall performance was strong with India and Southeast Asia standing out. Robust investment in India is expected to continue supporting its 2047 vision for long-term growth in the -- growth in ANZ remained softer due to the slowdown in transportation spending in Australia. However, there is general expectation that it will rebound driven by infrastructure projects tied to the 2032 Brisbane Olympics. China's performance was consistent with our expectations. Given the economic and geopolitical headwinds and represents only about 2% of total ARR.

And with that, Werner, over to you.

Thank you, Nicholas. We had a solid third quarter and are well positioned with respect to our financial outlook range for the full year. Total revenues for the third quarter were $376 million, up 12% year-over-year on a reported basis and 11% on a constant currency basis. Year-to-date, total revenues grew 11% and 10% on a reported and constant currency basis, respectively.

Our mainstay subscription revenues grew 14% year-over-year for the quarter in reported and 12% in constant currency. And for year-to-date, Subscription revenues grew 12% on a reported and constant-currency basis. Subscription revenues represent 92% of total revenues, up 2 percentage points from the same quarter last year reflecting improvements in the overall quality of our revenues visibility, growth consistency and margin contribution.

Our E365 and SMB initiatives continue to be solid contributors. Perpetual license revenues for the quarter were $11 million, essentially flat compared to the prior year. Perpetual license sales make up only 3% of total revenues and will remain small relative to our recurring revenues.

Our less predictable professional services revenues declined 2% for the quarter in reported and 3% in constant currency and now represents 5% of total revenues. We currently expect that our professional services revenues will remain at current levels for the remainder of the year. Hence, this would be for the full year, about $5 million less than we had originally planned. It is still the case that the largest portion of these nonrecurring services relate to IBM maximum implementation and upgrade work.

Our last 12 months recurring revenues, which include subscriptions and a small amount of recurring services increased by 13% year-over-year in reported and in constant currency and represent 92% of our last 12 months' total revenues, up 1 percentage point year-over-year.

Our last 12 months constant currency account retention rate remained at 99%, and our constant currency net retention rate rounded down to 109%, led in magnitude by accretion within our consumption-based E365 commercial model. We ended Q3 with ARR of $1.45 billion at quarter end spot rates. On a constant currency basis, our year-over-year ARR growth rate was 10.5%, consistent with our seasonality expectations for the year which included the favorable impact from the onboarding of our Cesium acquisition in '24 Q3 dropping off this quarter.

Excluding China, our year-over-year constant currency ARR growth rate was 11%, with China being 2% of our total ARR. On a quarterly sequential basis, our constant currency ARR growth rate was 2.2%, below our '24 Q3 sequential growth rate of 3.2%. And impacted by the timing of programmatic acquisitions and asset analytics deals.

With regards to seasonality, we expect '25 Q4 to have higher year-over-year ARR growth compared to '25 Q3 due to the timing of potential acquisitions and anticipated asset analytics deals. Our GAAP operating income was $84 million for the third quarter and $284 million year-to-date. I have previously explained the impact on our GAAP operating results from deferred compensation plan liability revaluations and acquisition expenses.

Moving on to adjusted operating income less stock-based compensation expense our primary profitability and margin performance measure. Adjusted operating income less SBC expense was $104 million for the quarter, up 16% year-over-year with a margin of 27.7%, up 100 basis points. Year-to-date adjusted operating income less SBC expense was $335 million, up 13% and with a margin of 30.2%, up 60 basis points.

Our margin performance for Q3 and year-to-date has been strong, and we remain confident about delivering our full year adjusted operating income less SBC target margin of approximately 28.5%, representing an annual margin improvement of 100 basis points. As a reminder, our OpEx seasonality is always more heavily weighted towards the second half with our annual raises occurring as of April each year, and our larger promotional and event-related costs also concentrated in the second half of the year and particularly Q4.

Further, our OpEx seasonality in 2024 was impacted from head cost run rate savings from our '23 Q4 strategic realignment, which benefited the first half of 2024 and shifted some of our run rate and discretionary investments into the second half of 2024. And particularly Q4 2024.

We therefore expect more than 100 basis points of margin improvement for the fourth quarter of 2025 when compared to 2024. Our free cash flow was $111 million for the quarter and $384 million year-to-date. This is generally consistent with our expectation based on our seasonality of collections and expenditures as well as the timing of cash tax payments, which are more concentrated in the fourth quarter. We are on target to meet our full year free cash flow outlook of $430 million to $470 million.

With regards to capital allocation, along with providing sufficiently for our growth initiatives. Year-to-date, we deployed free cash flow as follows: $135 million fully paying down our senior debt. $93 million in effective share repurchases to offset dilution from stock-based compensation, $10 million in convertible senior note repurchases and $64 million on dividends.

With our senior debt being fully paid down, our net debt leverage, including all of our 2026 and 2027 convertible modest was 2.2x adjusted EBITDA down from 2.9x at the end of 2024. Our strong balance sheet and projected free cash flow generation will sufficiently fund our dividend, share repurchases and growth initiatives, including potential programmatic acquisitions.

Our 5-year senior secured credit agreement day from October 2024 and provides a current undrawn $1.3 billion revolving credit facility. This provides sufficient flexibility to address the January 2026 maturity of $678 million in outstanding convertible debt while keeping our cash interest thereafter at about the same magnitude as in the recent past. Interest rates on our debt are protected through very low coupons on our convertible notes and very favorable terms of our $200 million interest rate swap expiring in 2030.

And finally, with only 1 quarter remaining, our performance for the first 9 months gives us confidence in our ability to achieve our annual financial targets. I already provided incremental color on our fourth quarter expectations for ARR, adjusted operating income by stock-based compensation margin and free cash flow.

With regards to total revenues, 2025 to date reflects a continued shift in mix from professional services revenues to subscription revenues, improving our overall quality of revenues and margin contribution. With regards to foreign exchange rates, for the third quarter, the U.S. dollar has weakened relative to the exchange rates assumed in our 2025 annual financial outlook resulting in approximately $10 million of incremental revenues from currency and a total favorable impact for the first 9 months of approximately $18 million.

Based on recent rates where the U.S. dollar has weakened relative to our outlook rates. If end of globe exchange rates would prevail throughout the remainder of the year our fourth quarter GAAP revenues will be positively impacted by approximately $8 million relative to the exchange rates assumed in our 2025 financial outlook.

And with that, we are ready for Q&A. Over to Eric to moderate. Thank you.

Thanks, Werner. [Operator Instructions] First question will come from Joe Vruwink from Robert W. Baird.

Maybe can you go into a bit more detail on the opportunity for better ARR growth in 4Q? That's a big renewal period, but also asset analytics opportunities. And just on the point about renewals. So to the point, Greg, you were making at the starts, how Bentley applications are called a few years from now, it could look a lot different than how they currently are utilized. How does that get encapsulated with an enterprise customer that is willing to engage with you over a multiyear time frame? And are you appropriately monetizing the full potential with kind of the ceiling floor structure you have been using around these consumption arrangements?

Well, I'll say to your last question, Joe, that we only monetize the actual consumption. It just happens to be bounded by a floor and ceiling potentially. And we are not yet monetizing API consumption, for instance, even though some of it is occurring.

What -- in the course of a renewal with an enterprise account for E365, they tend to prefer to get visibility into their spending in the out years as well, and it continues to be the case that we wind up on average, negotiating that each year of that renewal the floor and ceiling escalate by about 10%.

And I think they're aware because you see these enterprise accounts are the ones responding to the survey about spending on AI and expectations about AI. They know that their mix of consumption and modes of consumption will change over that period of time, but they are comfortable with expecting to spend the low double-digit amount more with us and no doubt with others each year, and we're satisfied to -- but we, likewise, now there's going to be volatility in the components of the mix.

But when you put it all together in an enterprise agreement, you've heard my take on that, which is to be confident that while the mix will change, the magnitude will reflect the increasing value, especially from AI.

Nicholas, as to the fourth quarter, indeed, it is a strong renewal quarter, and our expectations are comported with that.

Right. Well, first of all, Q3 ARR growth was exactly what we were expecting. And what we're expecting for Q4 is ARR growth year-over-year to be stronger than Part of that is the renewals as you mentioned. And then how much better it will be than Q3 will depend on potentially M&A or some of the big asset analytics opportunities that we pursue.

But it would be better than Q3 in any case because of the magnitude of renewals that occur in fourth quarter. So they are all layers that give us confidence in fourth quarter and of course, therefore, in the outlook for the year.

The next question comes from Jason Celino from KeyBanc.

Thanks. I wanted to ask about the government shutdown. I recognize that in your prepared remarks, you said it's had a minimal impact. Maybe can you just elaborate on what you're seeing or not seeing and why it's been so limited?

Yes. To date, we have seen indeed a minimal impact. First of all, our direct revenue from the U.S. federal government is less than 1%. And indirectly for the projects that were already awarded IIJA funding, while the funding continued to flow during the shutdown. And that's very much because of the way IIJA was structured.

Now depending on how long this shutdown is going to go. It could be that very much at the margin when we get to renewals with some accounts. And you may recall that renewals are based not just on past performance, how much have been consumed in the past year, but also I'm not expecting to consume in the next year, it could be at the margin the consumption expectations going forward may be impacted, yes. But this is super early to say. And really depend on how the shutdown goes.

Yes, hopefully, it ends soon. So yes, we'll see.

Other things kind of indirectly are gunned up also, and we hopefully, not for much longer, things like the permitting reform we keep expecting and some other functions of the federal government that don't have to do with using software, but changing policies and so forth that are also on hold. We'd like to see the shutdown and sooner rather than later, and we think that's the general expectation now.

Thanks, Jason. The next question comes from Matt Hedberg from RBC.

A lot of the bigger frontier model builders are noting that access to power is the biggest bottleneck for compute capacity today. And I guess, Nicholas, you noted both PLS and Seequent is set to benefit from this longer term, which is sort of in line with our view. It's also nice to see. I think although permitting reform takes time, Greg, you said it's sort of like there's -- it just takes time. It sounds like EMEA permitting reform is accelerating a bit.

I guess my question is realizing these projects take time and permitting reform takes time as well. Are you starting to see any sort of early benefit from early sort of like discussion with customers around this activity? And how should we think about sort of like medium to long term, this desire for more power to positively impact ARR growth?

Well, first, despite permitting reform still to come, both PLS and Seequent remain strong growth engines for the company, right? And both businesses are still growing faster than the company.

We have seen now in the U.S., for example, some execution of some mining projects through the -- it's called the [ FAST-41 ] process. For minerals that are strategic imports to the U.S. economy like lithium or copper, there is a similar bill that was passed in Canada. So we think this is I mean, this is happening basically for mining. There's already a lot of movement in order to accelerate permitting or accelerate certain projects.

But for the electric grid, this is still to come, and we've seen some very encouraging signs in the past few months in the U.S. I think there's a clear realization here that we must expand the electric grid. And there's a lot of effort, a lot of activities going into strengthening the existing grid, making it more robust, but we actually need to expand it in order to cook with higher demand that is for electricity and a lot of that coming from data centers by the way. So you can see it as a growing tailwind for us.

Thanks, Matt. The next question comes from Dylan Becker from William Blair.

It's [indiscernible] on for Dylan. I just wanted to double-click on to your AI innovation road map and how you're working with your customer base to build that out, maybe how that played into Cloud Connect and really what you're focusing on and how you're prioritizing the different opportunities?

Well, first of all, what was remarkable when we looked at the submissions for the -- this year's Going Digital Awards was seeing how much our users are investing in AI sales. And that was a big part of the update that Greg provided with the AEC adviser survey, and we can see how much is invested in AI.

And as always, we're using the Going Digital Award submissions as a bit our own survey of what's going on with the most advanced infrastructure organizations out there in leveraging digital to drive the productivity. And it points to our core applications -- engine applications, playing a new role going forward, not just here to empower infrastructure engineers, initial engineers, but actually to start to interact with AI agents. And we think this is a fantastic use opportunity.

Now we've seen a net acceleration of use cases where our own applications are being used in conjunction with AI that is being developed by our users. If we just look at and reflect on the past couple of years, and we've announced a co-innovation initiative in order to engage with our users to partner with them to discuss how can we evolve or engine applications, technically and also how can we evolve the commercial model around these applications so that we can support those workflows going for -- better support as workflows going forward. We're rurally excited about that, right?

But in there is a lot of investments on our side for the -- for our own AI capabilities that we're delivering to our users. Here, we make sure that, first of all, all of our coorganization, all of our user-facing teams have a deep empathy, a deep understanding of the needs of the users, the accounts that we serve. We understand where potential opportunities to drive more productivity through AI, for example.

And then we involve represented users along the way in helping us prioritize which use cases that we're going to go after first with AI, we involve them during the development of those capabilities. We involve them with beta software, what we call early access. We involve them, of course, the limited availability to make sure that the product can scale to the broader market. and so on and so forth.

So we have constant touch points all the way from the very beginning of the exploration -- can resolve with the eye to making sure that the software can scale. We have involvement all the way with representative users.

Thanks, [ Pete ]. The next question comes from Kristen Owen from Oppenheimer.

So I wanted to ask you about labor availability, not just here in the U.S. but globally in the construction and infrastructure trades. Obviously, AI can't actually build infrastructure. So I'm wondering if you're starting to see that meaningfully impact any of your engineering from customers what sort of impact these labor challenges are having maybe on project delivery times, budgets and then add on this piece of willingness to invest in technology to help with some of those productivity challenges?

Kristen, I think the biggest picture is that everyone has long expected the engineering services firm. So that's half of our business. And they work for the other half of our business, the owner operators, everyone has expected the way they work to change from a time and material billing hours to paying for value and being therefore, having a platform to incent and reward for instance, these AI investments.

The biggest picture, I think, is that the ongoing engineering resource constraints are influencing that now happening in favor of AI investments and expectations and changes in the commercial model. And the opportunity for us is to be shoulder-to-shoulder alongside those engineering terms, we want to help be their arms merchant providing them the, for instance, asset analytics, cloud services that they will rebrand and bundle with their engineering analytics and their own data and AI models.

But where they won't need to get into providing the cloud services, the things we can do together with Google and then adding our asset analytics are. So changing the commercial model is the -- accelerating that because everyone has expected it to occur, and it's been slow. I think that finally is being catalyzed now, and that's the biggest impact.

The next question comes from Alexei Gogolev from JPMorgan.

I wanted to ask you to maybe give us a brief update of how the partnership with Google is going? Have there been any incremental customer conversations on the back of this partnership? And what does that mean for your asset analytics opportunity?

One of the updates we gave at YII was how we are integrating Google Geo data across the portfolio. It starts with orange applications, MicroStation, the new version MicroStation, include Cesium for 3D geospatial visitation, but through Cesium is actually integrating 3D photorealistic cars from Google.

And I mentioned the launch of Bentley Infrastructure Cloud Connect the user experience of an infrastructure cloud is also powered by Cesium and also powered by 3D photoreceptors of Google. So that integration is going very well. and we're expanding really across our full portfolio. And we are quite excited also about the opportunity with Google when it comes to asset analytics.

Google is a source of data that can be analyzed in order to better understand the current conditions of infrastructure assets and their full context. And you may recall that we've announced a couple of months ago, a deeper partnership with Google from that standpoint, where we'll be processing Google Sreetview imagery to understand basically the inventory of assets out there and be able to do a before and after comparison on what's going on with this infrastructure asset when we compare with Dashcam data that we're processing through our own road monitoring solution, right? And that's just 1 example of so many other use cases that we're discussing with Google, where we can empower deeper analytics about existing assets.

Thanks. Next question comes from Clarke Jeffries from Piper Sandler.

I wanted to ask just a little bit of a follow-up to the discussion around the appetite of AI spend with your customers seems like a lot of this is survey work and sort of perspective on where they'll go. But I wanted to ask today, are you seeing proactive RFPs from these customers around AI capabilities? Or is it too early? And sort of do you imagine there being a discrete sales approach around AI functionality? Or do you feel like this will be very organic within your kind of existing sales motion?

Also on the former, it is still too early for the market you asked specifically for AI capacities for specific use cases. It's still too early. However, indirectly, we do see infrastructure organization insisting that the -- that it is as easy as possible for them to access data that is being created or managed through software coming from providers so that they can use it for their own AI purposes.

And by the way, so this is much point direct, yes, but they are aware that software providers are developing capabilities, and we see them clear and clear about we want to make sure that when you do this as a software provider, you don't use our data without our explicit permission, right?

This is very top of mind right now. And it's a part of the conversation we're talking about Bentley Cloud Connect, for example, we reinforce our commitment to data stewardship. We make it very clear that the data of our users is their data always. We don't use it to train our own AI unless they explicitly authorize us to do so, yes. And that is a clear differentiator versus other providers who are maybe -- well, just less clear on that topic. But in general, are still very early stage when it comes to requirements for AI specific capabilities.

And because it's still very early stage, we don't see a need right now to have a different go-to-market approach when it comes to position is AI capacities. If you look back at what we basically announced in terms of our AI capabilities, sometimes it's the next generation of an existing application like OpenSite+ is the next generation of OpenSite+ an negation of substation the same way that we've gone to accounts position opens the original OpenSite solution, the -- substation solution, and we're going to continue to do the same even if -- the new one is powered by AI.

Then we're also introducing a lot of new air capabilities are inabilities to existing applications. So same thing there's no need in order to do a different go-to-market. I said antics is different. Here, we've been always very clear that we want to go both direct and indirect, right? So I said we are going after owner operators and the firms that serve them in order to position those caplities but we definitely welcome all the organization to take our capabilities and offer them as part of their own offering for asset monitoring, asset maintenance, asset management.

Next question comes from Jay Vleeschhouwer from Griffin Securities.

Nicholas, I'd like to ask about something we talked about at the conference 3 weeks ago in Amsterdam having to do with your product development. Specifically, how have you evolved or how you think you might still need to evolve your product development, management or operations in light of all that you mean to do across the portfolio? Do you think that you can or should, for example, impress the time between beta and GA, something we talked about a few weeks ago.

And in light of what Greg talked about earlier with regard to your new consumption model, how make those new techniques of consumables possibly tie back into your product development process or product release timing?

Right. When it comes to our own product development process, we got an earlier question here on how much we're involving users, how much are we involving accounts. And we're very keen to continue to do that and do that even more, right?

When we release our software, we want to do it in a very iterative fashion. We don't have necessarily top-down target on you must go from early access beta to limited verity in that time frame and you must go from entity to generate in that time frame. It will really depend on the feedback we're getting from representative users, especially when it comes to very new capability, we're sometimes creating new potentially, we're creating new markets, when it comes to asset -- et cetera, we want to make sure that we get it right. before we just -- we pushed too hard, right?

So it's important for us that we keep -- it is a very close feedback loop with representative users, and we trade as often as necessary in order to get the software right and then ready to scale. And now we are embracing AI as well to improve our productivity. And also the majority of developers now are working with AI tools every day. to be more productive, whether it's for coding assistance or even generation of some parts of the code for mundane and mundane functions, right? And we're quite pleased to see the level of embraced by our developers, not necessary to own really coming from them to use AI capabilities. And it remains very good and energy for how we're seeing AI playing out for infrastructure engineers, not AI replacing infrastructure engineers, but AI making infrastructure engineers more product, AI being more of a competitive, if you thus the name, by the way, of our own AI system.

Jay, I'll say that you and I share a long background going back to when our desktop products were a platform for specialized applications developed, including by our accounts. And we had special teams that worked with the accounts, the developers within the accounts to help support their particularizing our existing applications for their own purposes.

That's kind of gotten extent by now. Individual organizations don't develop their own particular software for this. But what we saw in the AI surveys, and we saw under Going Digital Award submissions are a lot of investment by the enterprises, the AEC firms, larger ones in their own agent environment. And the APIs that we'll create because we have to move engines to the cloud and open them up and so forth, will be another way of working with developers. The developers will turn out to be the developers in the large enterprises.

And I can see that being a different kind of go-to-market incremental orthagonal approach for the future. echoing back to what we've done in the past, where we love our role as a platform provider, especially.

Thanks, Jay. Next question comes from Taylor McGinnis from UBS.

Maybe just on the financials. So if I adjust for the lapping of the acquisition, it still looks like net new ARR was down a bit year-over-year on a constant currency basis. So I know you guys said that, that was in line with expectations, but maybe you can just unpack the drivers behind that?

And as we look into 4Q, I think to get to the upper end of your guys' guidance range, it implies a big step-up in net new error. I know you mentioned M&A and some of these asset analytics deals potentially being needle moving there. So when you think about the size of M&A that you guys are contemplating or how big some of these asset analytics deals could be, could you just provide a little bit more color there?

Well, okay, I'll just jump in on asset analytics because you've heard me say, we have such a big dependable flywheel. But the only thing that's a volatile in what we do is the asset analytics business because we're looking at landing 7- and 8-figure deals. And they -- it isn't yet at critical mass, I think you could say that. We believe it will become so soon.

But on the margin, it does make these differences in which quarter those deals fall. And of course, speaking of frame of reference, our business used to be like that back when the software business was an upfront license business and so forth. But for us, it makes this difference on the margin, but the margin is what is significant when we're comparing 1 quarter to another at the level of when was 10.5%, 10.9%, and so forth, it's to do with these asset analytics teams.

Yes. Maybe I'd add like in asset analytics, like Q2 and Q3 last year was particularly strong. And the opportunities for bigger deals more towards the end of the year in 2025. And on the M&A side, so just to say like we don't need M&A to be within the outlook range for our ARR. There are a number of transactions that we are working on. We expect that we close at least one by the end of the year. Over the last few years, our contribution to constant currency ARR growth through M&A was between 40 and 70 basis points.

And we do expect for this year that we are roughly within that range. Again, and our Q3 year-over-year ARR growth rate that we're showing is purely organic. So there's no benefit from M&A at all. So 10.5% purely organic. And without China, it will be 11% purely organic.

Thanks, Taylor. Next question comes from Siti Panigrahi from Mizuho.

I want to ask you about macro. If you think about last year, there was so much uncertainty election going on interested high. How do you view the macro now on this environment right now heading to 2026?

And Werner, anything that we -- any puts and takes that we should think about 2026?

Well, I'll start on the macro. It remains robust. The backdrop is the same. The end markets are strong. There's never been more demand for better, more within infrastructure. around the world. The only exception we've talked about for a long time now in China?

And then I mentioned in my prepared remarks, briefly Australia. In Australia, it's a bit of a crosswind you have less investments in the transportation infrastructure that we've seen in the past few years. But on the other hand, you have more investments going into mining. So as we get into -- we look start to look into 2026. We are not expecting a change of the overall demand environment. We expect the demand environment overall to remain robust.

I'll say that a difference from a year ago in the world, if we step back, is this notion, unfortunately, that each country needs to be self-sufficient and it's in its resources and requires infrastructure investment, if you like, given some redundant infrastructure investment to do that.

And then the other factor, for instance, the COP conference this year, the theme is on adaptation and adaptation as part of resilience is the work of civil and structural and geotechnical engineers. And it's just understood we need to get on with that ever more. those are changes that may be resulted from politics, but they end up adding to the demand for the work of infrastructure engineers. And again, there aren't enough of them without Going Digital.

Next question comes from Guy Hardwick from Barclays.

Just a quick one for me. So last month, they were speculated in the press of a merger between the #2 E&C firm globally and the #4 player globally. I was just wondering, consolidation amongst your larger E&C customers, what are the kind of positive and negative implications potentially for Bentley?

Well, some of that has taken place in the past and has not been to any disadvantage. The in other types of, if you like, design software, it might be R&D functions that would be consolidated.

The way that our software is used by the engineering and construction firms is in their throughput of production. It is a means of producing their product is their factory floor, if you like, and a combination makes them larger, but need no less software. And we sort of tend to be the choice for larger firms.

The consolidation, I think, has ultimately benefited us because of the type of technical platform cooperation that we're salivating now to do with expanding our APIs for AI to have our analytics and simulation engines be available for the development to be used to provide the engineering acquisition in context in AI developments that the larger firms, as Nicholas pointed out, are more investing in. It's going back to us, I was saying with Jay, this notion of being technically shoulder to shoulder as well as commercially shoulder to shoulder. I think that benefits from consolidation.

The next question comes from Koji Ikeda from BofA.

Listening to the call in the commentary, lots of commentary on external AI opportunities out there for Bentley. But I wanted to ask about an update on how you guys are internally using AI to drive productivity gains and sales, R&D, G&A and longer term, what could the internal use of AI mean for margin expansion for Bentley?

Yes. Thanks for the question. I didn't touch on it when it comes to your own internal use of AI for product development. But you're right, we're actually expanding the use AI of across business functions. And we've seen some quality improvements in lead nurturing, for example or user support.

So we see AI definitely as a way of making our existing colleagues more productive. And therefore, it will help to increase both the top line and the bottom line. Yes, that's our expectation going forward.

Our last question comes from Joshua Tilton from Wolfe Research.

Can you hear me?

Yes.

I've been jumping around this morning, so I apologize if it's been asked, but I think it's very important.

Greg, you always stress how predictable this business is. And I think that's what people really love and enjoy about it or at least it's one of the things I think they do. Last quarter, you had already told us that you expected this to be the low point of ARR growth for the year.

I guess my question is, was that in line with your expectations? Or did it come in even below what you guys thought? And the reason I'm asking is because, should we just view this as right down the fairway with your expectations and continued confidence into Q4? Or did things maybe trend a little bit worse than you were expecting even below that low point you kind of guided us to last quarter.

Well, I think it's the former, but Werner, I think it's worth wrapping up with a summary of the factors and how that's different for Q4.

Can just repeat the question, sorry.

I was just asking, you told us that this was going to be the low point for the year. And I think we're just trying to understand was that low point in line with your expectations, and we're just before as we were 90 days ago when you told us this was going to be the low point or was this low point worse than you were expecting and that we should adjust your expectations for Q4?

Understand. I understood, sorry. So I think we are exactly where we expected to be for Q3. It's clear that Q4 is a big quarter for us, like most of the renewals are in Q4 or like not most of them, but a very significant amount of our annual contract renewals are in Q4.

We see the pipeline is as we expected in our outlook. And we will focus on strong execution as we did year-to-date. And then we -- as we said, like we have the opportunities within asset analytics and programmatic acquisitions that makes us confident that Q3 will be the lowest point and we are going up to -- from a year, if you will. We feel the same as we talk a quarter ago. So we are on target.

I don't know whether you called it Joshua, but Werner quantified, the year-to-date contribution from programmatic acquisitions in our year-over-year ARR growth is 0 this year. And it's generally 40 or 70 basis points and we actually may wind up there because we continue to strategically prioritize asset acquisition opportunities.

And the asset and acquisition opportunities, just to go back to that, are the lumpy deals. And back when lumpy deals were part of our business, which has been a long time ago, we remember they usually occur in Q4 and this year doesn't seem different, even though as we got started with asset analytics, last year, we had some big deals in Q2 and Q3.

So again, it is a big reliable flywheel, but it has on the margin these changes. And I'm gritting because I think those are the right things for us to be doing. I said analytics is the ground floor of a huge opportunity to AI-enabled.

And even though it's going to be a bit of a nuisance, its volatile nature ultimately, we'll spread it all out as we gain critical mass and escape velocity there, as I say, and I feel that's coming closer strategically.

Makes sense. Very helpful. Maybe just 1 last 1 before you kick me off. You guys had the year in Infrastructure Conference, lots coming out of it. We also had Autodesk University like a quarter ago. So announcements across the industry. I guess if there is 1 announcement that you think is going to be the most needle moving for the business that investors should be paying attention to? Like what would you call out from the year Infrastructure Conference?

I would say, short-term connect -- connect. Yes. You heard new formation layer for Bentley Infrastructure Cloud, bringing a lot of capabilities that used to be and the different surprise systems that we brought together under the umbrella of Bentley Infrastructure Cloud. And basically where data is being federated in order to be used for AI purposes.

And I would just say, stepping back, a contrast among these announcements. You have other vendors out there whose products have always been separate and there's been no way to get data from 1 to the other.

Of course, AI models enable them to say, okay, if you pay us now, will actually help you get data from 1 application to another or from our standpoint, it's all been integrated a common schema, we're way ahead on that point. It's not a matter of monetization. It's a matter of improving the form factor to make it even easier to use and even easier to use with AI agents and API consumption.

I'd say, AI and put the P in the middle. That's how we want to be a platform vendor to fit that into our existing enterprise accounts. And then of course, a different go-to-market motion for AI for the SMB firms, but the -- what I think Nicholas says that the connect is important because it brings this down to the level of every user so that it's not -- so that it's intuitive and immersive and geospatial and new.

But the back end of how things are integrated together, we're not inventing now we're leveraging now.

Thanks. That concludes our call today. We thank you for your interest in time and Bentley Systems. Please feel free to reach out to Investor Relations with further questions and follow-up, and we look forward to updating you on our performance in coming quarters. Thanks a lot.