EPAM Systems, Inc. (EPAM) Earnings Call Transcript & Summary
August 6, 2026
Earnings Call Speaker Segments
Mike Rowshandel
executiveHello, everybody, and thanks for being with us for the Q2 2026 earnings announcement. As the operator said a moment ago, my name is Mike Rowshandel, and I serve as Head of Investor Relations. We hope you've already had a chance to look over the earnings release that came out earlier today. If you haven't, you can find copies at epam.com under the investors section. Joining me on this call are Balazs Fejes, our CEO and President, and Jason Peterson, our Chief Financial Officer. I want to remind everyone listening that certain comments we make today might include forward-looking statements. Those statements carry risks and uncertainties, which are described in the company's earnings release and SEC filings. Also, any references to reported results that are non-GAAP measures have been reconciled to the comparable GAAP measures, and you can find them in our quarterly earnings materials in the investors section of our website; so with that, I'll hand the call over to FP.
Balazs Fejes
executiveMike, thank you, and good morning, everyone; it's a pleasure to be here with you once more. Over the past quarter since we last spoke, I've spent my time with customers, partners, and our own colleagues across our delivery centers worldwide, and one thing keeps becoming clearer to me. Every part of what we do is being transformed by AI, and at the same time enterprises are facing more complexity than ever. The gap between trying AI out, truly embracing it, and getting the most from it — that's where EPAM's opportunity lies. Revenue growth for the second quarter came in at the top of the range we had forecast, and we also saw continued improvement in profitability along with strong adjusted earnings per share. Our revenues that are purely AI native accelerated, exceeding $160 million in the quarter. We're still seeing solid operating momentum, and on this call we want to be straightforward about where growth is falling short—particularly in large areas of the North American business—and what actions we've already begun in response; the plan introduced at March's investor day is being executed, and this quarter gives additional evidence that our direction is the right one. Conditions over the last few months have been turbulent and unpredictable, both broadly and within our industry specifically, and that instability reflects the extent to which AI and related technology remain disruptive. Our results for this quarter confirm the message we've been sharing. Artificial intelligence is driving the need for niche skills and new kinds of abilities, and to capture these emerging opportunities we're carrying out three strategic pillars. Starting with pillar one, being a leader in AI native software engineering, we are transforming our software engineering approach from within, moving customers beyond the conventional SDLC toward completely integrated agentic enterprise harnesses, real-time data modernization, cyber resilience, and token cost engineering. At present, advanced instruments like AI/Run, DIAL, and MF Lens are handling sophisticated use cases that have been deployed by thousands of specially trained EPAM engineers spanning hundreds of client engagements. During every client meeting I had this quarter, one question kept coming up: "How do we get value out of AI and see a positive return on investment?" What I tell them is that there's no easy answer here, because AI doesn't make things less complicated. It actually piles on more complexity — spanning talent, architecture, process, governance, and models. Coding may become automated, but engineering won't. As AI improves at producing code, what matters more is the last mile of solution engineering and getting deployments to succeed. Each time we get an AI solution live and working, it opens up fresh scenarios and broader scope for the client. That's precisely the reason the groundwork clients have been doing for years—modernizing systems, building out data engineering, and clearing away technical debt—now represents our single largest opening, and it's one we can finally go after thanks to AI. Still, we recognize this cycle remains at an early stage, and it doesn't move in a straight line. Certain initiatives are kicking off, others are being halted, and a number are shifting into different engagement models, which reshuffles the mix and reshapes the market. On the second pillar—making EPAM a full stack AI-native organization—we keep picking up speed and widening our strategic partnership. Those partnerships, along with others, sit on badges. That's the way we assemble a multimodal bench that's already supplying named pipeline along with plenty of our bigger multiyear deals. During this quarter, we became part of the OpenAI Partner Network, joining as an OpenAI advanced partner with a route toward elite status. Jointly, we're developing forward-deployed engineering, cyber resilience, and customer experience capabilities atop OpenAI's frontier models, with a first-year commitment to certify over 5,000 OpenAI consultants and to train more than 10,000 EPAM specialists. Beyond building that expert team, we're growing into security, managed services, and solutions tailored to specific industries across our core regions. Working with Google, we've gotten more than 2,000 of our people certified out of a Q3 goal of 5,000 through the Gemini Enterprise Certified Partner Specialist program, which strengthens what we can do within Google's Gemini enterprise ecosystem for creating, operating, and overseeing multi-step AI agents. On top of that sits our years-long GCP partnership, which includes upwards of 2,000 certified Google Cloud practitioners, agentic offerings on the GCP marketplace, and award-winning solutions. Thanks to Anthropic, we're now counted among the five highest-ranked globally certified partners, with over 5,700 certified engineers, which already puts us past the 5,000-by-Q3 target we had set. Our goal is to have more than 10,000 Claude certified architects by the time this year ends, and already over 50% of the people in our delivery organization have finished the Anthropic academic coursework. A committed team of 250 forward-deployed engineering black belts stands behind all of that. It's worth noting that the practice has moved beyond simply enabling teams — it now reaches into security, into specific verticals, and into key accounts. Finally, regarding the third pillar, our go-to-market transformation, we're using our AI-native structure to launch fresh go-to-market motions, and right now North America is where we're placing particular emphasis. Last quarter, we began a planned, multi-quarter overhaul of how we sell — creating one consistent method for deciding which major accounts matter most and how to go after them, putting tighter processes around tracking potential new customers, and putting money into sharpening our team's selling skills through company-wide coaching. This is money spent now to strengthen how we sell, and it reflects the return to basics in how we run everything. Moving on, here are a handful of Q2 points worth noting. On a reported basis, revenue climbed 4.5% compared with the same period a year earlier, and in organic constant currency terms it was up 3.4%. Growth occurred in four of our six verticals versus a year ago, with financial services along with life sciences healthcare setting the pace, and gains also came from emerging verticals as well as consumer goods and retail and travel. During the quarter, both software and high tech, as well as business information and media, experienced a downturn. For business information and media, the drop stemmed from wrapping up a number of client projects. In software and high tech, we saw reductions in project scope focused on non-AI services, and those reductions were larger than the gains we achieved in AI, cloud, and cybersecurity work inside that same vertical. Looking at different regions, EMEA continued to set the pace for our growth, putting up a strong double-digit result, whereas the Americas turned in much weaker growth by comparison. Now, shifting to what we're seeing on the demand side: when it comes to the broader macro picture, how clients feel and how they're handling budgets are basically in the same range they were in the prior quarter. Conditions haven't shifted in any meaningful way, either for better or worse, and with conflict continuing in the Middle East, we're still watching decisions take a long time to get made. Even so, there remain certain genuinely strong spots. I'd like to walk through a few details. Revenue growth in EMEA remains robust, powered by a lively pipeline that comes from the proactive go-to-market sales motions we've already put in place. Once again this quarter, financial services stood out as our fastest growing vertical, with gains delivered across EMEA as well as the Americas. One big source of growth has been putting AI to work on customer use cases — mainframe modernization stands out here, since it relies on EPAM IP to reverse engineer, automate, and rebuild through newly added forward-deployed capabilities. In the quarter just reported, life sciences and healthcare came in as our number two vertical by growth rate, building steam over the past two quarters, driven by pharma R&D and clinical trials combined with AI, as well as continued traction in med tech products and services. Turning to energy, the amount of business we hold is much larger than it was a year ago, largely thanks to expanding what we deliver to clients we already serve, plus revenue brought in by new logos. Though our historical focus was upstream, we're now growing our book of business to include midstream, downstream, and data center work in this vertical. Now for a few of the offsetting factors. I'll be straightforward.
Balazs Fejes
executiveNorth America isn't expanding at a sufficient pace. Our current projection is that it will perform under our expectations during the back half of the year. Judging by where the business stands right now, let me be direct: this isn't a situation where we're just waiting for macro conditions to improve. The responsibility is ours. Two factors are creating these conditions. First, there's a real change in what clients in North America are purchasing. Demand is shifting away from tax-based services such as manual testing, user experience, JavaScript front-end engineering, and moving toward AI-led modernization. The shift is happening faster than the substitute work can grow, so that leaves a gap in growth we have to shut even more rapidly. On a different note, this quarter brought a decline in software and high tech, largely tied to projects winding down. Even though the client ties at the core remain strong, the scheduling is creating a weight on that section of the portfolio. Also, our regional approach to going to market has not been delivering at the level it needs to. That part is fully ours to fix. The commercial overhaul spanning multiple quarters that I brought up before is exactly what will address it, creating momentum for forward selling as we move through the coming quarters. At this point, we'd rather be upfront about what to expect than have you hold out for a turnaround we haven't proven we can deliver. Jason is going to walk you through how that plays out in the figures. Now, on to the pipeline for new large deals. Things are moving in a positive direction here; every one of these AI-driven opportunities involves clients we already have, rather than new logos. They keep advancing through our pipeline on an active basis. Each one centers on AI, more precisely agentic managed services along with application maintenance. Let me be clear: as of now, not a single one has been signed. We're intentionally holding back from counting on them prematurely. The upside is genuine. Of everything this quarter, it's among the developments that gives us the most encouragement. These deals stand out because of how they're put together, how big they are, and the fact that they span multiple years. Through our AI-native strengths, we're going after pieces of work at existing clients where EPAM traditionally hasn't had a seat, which lets us grow our footprint and our impact — it's incremental growth layered on top of the core business. Our attention has been on getting this done. That said, the usual procurement timeline stretches out further than what we typically see. Given how things look today, along with the odds of getting these signed and when the ramp would kick in, our expectation now is that a more meaningful revenue contribution begins in the first half of 2027 rather than the second half of 2026. Now on to AI. This quarter, our data business outpaced growth in the rest of the business. That core need is precisely what keeps supplying our pipeline of AI-native business, which reinforces our belief in the enormous, mostly untapped backlog sitting beneath AI, technical debt, legacy modernization, plus foundational data and cloud initiatives — all of which must take place before AI can deliver value to the enterprise. Yet although revenue from AI-native offerings continues to accumulate nicely, extending its run of quarters marked by double-digit sequential growth and now representing over 11% of what we do, moving it toward a larger share of the business is going to take patience. The constraint on us has nothing to do with our ability to execute. Our FD teams, the AI/Run platform, and our partners are able to take on far more than the backlog currently in front of us. What complex enterprises require for dependable momentum is an approach that carries AI past enablement into actual business transformation and results. Set beside where we stood 12 months ago, the strides we've made are genuine and significant. The sector as a whole remains fairly early in this journey, all while navigating fast-shifting and intricate industry headwinds. Allow me to walk through several client stories that demonstrate this. With a top-tier financial services wealth management organization, we're bringing AI into its modernization efforts and speeding up mainframe transformation by reverse engineering beyond 10 million lines of code, achieving 60% in time savings on reverse engineering work, all while embedding fresh forward deployed engineering skills into entirely new ways of engaging. With a multinational beverage manufacturer, we assembled a unified data platform that pulled scattered enterprise data together into a dependable AI foundation, which allowed them to scale and drive forward more than 80 data and AI initiatives. Now that it's live, that company has realized more than EUR 30 million in business operational impact across the last 12 months; with a global energy commodity business, EPAM assisted in shifting over 1,000 workloads onto AWS with no downtime experienced by users, which produced a 40% cut in infrastructure and operational costs alongside a 30% gain in operational efficiency. This effort unfolded during a delicate period following an acquisition, when that company required extracting a business-critical application from its legacy setup and merging hundreds of aging on-premises servers. Since our last conversation, we've been privileged to earn a number of notable leadership honors. At its 2026 awards, Databricks named EPAM Consulting and Systems Integrator AI Partner of the Year. That recognition points to how EPAM helps customers in many industries move AI from theory into real use, turning fragmented data into AI applications and agents built for production. In cloud security, EPAM took home the 2026 Fortress Cybersecurity Award for migrating a Swiss private bank's entire IT estate — hundreds of applications in all — onto Microsoft Azure while meeting strict FINMA mandates. Meanwhile, Gartner's emerging market quadrant for physical AI services lists EPAM as a specialist, a distinction that sets it apart from long-established competitors. And in IDC MarketScape's inaugural 2026 vendor assessment of worldwide life sciences R&D strategic consulting services, EPAM appears as a major player. Forrester named EPAM one of the 28 leading vendors in the customer experience strategy consulting services space. The Wall Street Journal also recognized EPAM as one of its best companies for the future. Acknowledgments like these continue to reflect the hard work and dedication of our teams around the globe, as well as our unwavering commitment to delivering tangible results and outcomes for those we serve. To sum up, our strong Q2 shows genuine progress on our multi-year strategy and our AI-focused goals. Despite the macro conditions we encountered, we came out ahead, made real strides across all 3 of our strategic pillars, and intentionally took first steps in the areas that needed them most, particularly North America. Our confidence stays firm in the long-term plan to turn into a worldwide frontrunner in AI transformation services, acting as a dependable AI accelerator and an ally for our customers. The momentum behind our AI native and AI foundational offerings keeps growing. As time goes on, those revenues will represent an ever-bigger portion of the whole business. This quarter, in our view, marks a strong move along that path. We recognize plainly that the back half is going to be uneven, especially in North America. Our belief in this strategy, in this team, and in the commercial transformation is strong. What lies ahead across the coming 2 quarters is to keep turning attention into outcomes with the very same discipline that brought us to this point. Finally, I'd like to express gratitude to everyone for your ongoing dedication, confidence, and backing. Jason, the floor is yours.
Jason Peterson
executiveThank you, FP, and good morning to all of you. During the second quarter, we posted robust results for the period, beating the upper bound of the outlook ranges we had provided for organic constant currency revenue growth, profitability, and EPS. Revenue came in at $1.415 billion, which represents growth of 4.5% versus the prior year. On an organic constant currency basis, revenue rose 3.4% against the second quarter of 2025. Both GAAP and non-GAAP gross margins improved compared with a year earlier and came in above what we had anticipated for the quarter; thanks to healthy revenue growth and better year-over-year profitability, GAAP income from operations increased by over 20%, while non-GAAP income from operations rose 14.7%. GAAP diluted EPS was up 26.3%, and non-GAAP diluted EPS climbed 22%. AI native and AI foundational revenues keep adding to our year-over-year growth. With AI native revenues exceeding $160 million during the quarter, we have now seen six straight quarters of sequential double-digit expansion. Turning to how our Q2 industry performance looked, we achieved year-over-year growth in most of our verticals. Financial services turned in strong growth and stood out as our fastest-growing vertical this quarter, rising 11.5% year-over-year, largely fueled by insurance and asset management clients throughout both EMEA and the Americas. We keep producing double-digit revenue growth in that vertical, with AI modernization programs contributing significantly. Consumer goods, retail, and travel posted 2.3% year-over-year growth, led notably by retail and consumer goods. Life sciences and healthcare turned in solid growth and ranked as our second-fastest-growing vertical this quarter, at 8% year-over-year. Growth in that vertical continues to come mainly from clients in life sciences and med tech. Importantly, year-over-year revenue growth in this vertical keeps accelerating. Software and high tech fell 1.3% year-over-year, largely because of the anticipated ramp down of a large client program, along with the shift in priorities that FP discussed earlier. Business information and media dropped 2.1% year-over-year, mainly due to several client projects wrapping up. Our emerging verticals achieved year-over-year growth of 4.9%, mainly on continued strength in energy and manufacturing. Looking at it geographically, the Americas, our biggest region, accounting for 57% of Q2 revenues, was up 0.5% year-over-year, as strong financial services revenue growth was substantially offset by declines in the software and high tech and business information and media verticals. EMEA, which made up 41% of our Q2 revenues, rose 10.9% year-over-year and 9.4% in constant currency, helped by strong financial services growth along with contributions from travel and consumer goods and energy. And finally, APAC, representing 2% of our revenues, was down 0.3% year-over-year. Last, during Q2, revenues from our top 20 clients rose 3.1% year-over-year, whereas revenues from clients beyond our top 20 were up 5.2%. Now, as we continue down the income statement, our GAAP gross margin for the quarter reached 30.4%, versus 28.8% in Q2 of last year. For the quarter, non-GAAP gross margin came in at 32%. That's up against 30.1% during the equivalent period 12 months earlier, which shows we can boost profitability even while we keep putting money into our AI capabilities and expanding our AI native revenues. GAAP SG&A represented 17.3% of revenues, versus 17.1% in Q2 of the prior year. Non-GAAP SG&A for Q2 2026 landed at 14.5% of revenue, up against 14.1% during the same stretch last year. GAAP income from operations reached $152 million, which works out to 10.8% of revenue. A year earlier in Q2, that figure stood at $126 million, or 9.3% of revenue, meaning growth of 20.4% year-over-year. Non-GAAP income from operations hit $233 million, equal to 16.4% of revenue. In Q2 of the previous year, it was $203 million, or 15% of revenue, delivering 14.7% growth year-over-year. Our GAAP effective tax rate, which takes in a greater amount of tax shortfalls tied to stock-based compensation, arrived at 26.7%, while our non-GAAP effective tax rate was 24%. On a GAAP basis, diluted earnings per share came to $1.97, versus $1.56 in Q2 of last year. That's a $0.41 rise year-over-year, which reflects growth of 26.3%. Non-GAAP diluted EPS reached $3.38, compared with $2.77 in Q2 of last year. That's a $0.61 gain year-over-year, reflecting growth of 22%. During Q2, diluted shares outstanding totaled roughly 52.3 million. Now on to our cash flow and balance sheet. Cash flow from operations in Q2 was negative $2 million, versus a positive $53 million during the same quarter of 2025. Higher variable compensation payments tied to 2025 performance, along with a rise in DSO during the quarter, weighed on Q2 cash flow. Free cash flow came in at negative $18 million, compared with positive free cash flow of $43 million during the same quarter last year. As of the close of the quarter, cash and cash equivalents stood at roughly $800 million. At the conclusion of Q2, DSO was 82 days, which compares with 76 days for Q1 2026 and 78 days for the same quarter last year. Over the quarter, we bought back roughly 1.3 million shares, including open market purchases of about 800,000 shares for $80 million. Roughly half a million shares came from the final settlement of our accelerated share repurchase that was paid during the first quarter. Since our share repurchase program began, we have returned about $1.6 billion in cash to shareholders to date. Let's move on to operational metrics. We closed Q2 with over 56,650 delivery professionals, which reflects total growth of 1.5% against Q2 2025. Our total head count when the quarter ended was over 62,850 employees. Utilization reached 78.3%, compared with 78.1% in Q2 of last year and 77% in Q1 2026. Now let's shift to guidance. Before I get into the specifics of our 2026 and Q3 outlook, I'd like to share some thoughts to help frame our guidance. Our Q2 results leave us feeling good, and the momentum we keep seeing in pure AI-native revenue supports that. That keeps us positioned to hit our target of $600 million from AI-native revenues in 2026. At the same time, we've managed to lift company profitability, and gross margin in particular, even as we keep putting money into our growing AI capabilities. Looking ahead, we now anticipate a deceleration in the rate at which our revenue grows during the second half of the year. Let me lay out plainly what sits behind this revised outlook. As FP said, North America gave us modest growth in Q2, and for the rest of the year we expect revenue growth from that region to stay very slow. Our view is that the weaker growth in North America is mostly something specific to EPAM, and it's an issue we're working to fix. Meanwhile, our emerging pipeline of larger opportunities keeps building, and the scale of that pipeline and the strides we're making in negotiations with clients continue to encourage us. Not one of those deals has closed. Because of that, we don't anticipate meaningful revenue from those deals adding to growth until early in 2027. For Q3, we now expect modest sequential growth, and going from Q3 into Q4, revenues should be flattish. Our revenue guidance for 2026 will be coming down.
Jason Peterson
executiveMeanwhile, given how profitably we performed during the first six months and a revised projection for the back half, our expectation now is that we'll land at the upper portion of the adjusted IFO range we previously gave of 15%-16%, and our guidance has been revised accordingly. Relative to 90 days earlier, nothing materially better or worse has emerged in the wider macro picture; we aren't building in any shift, whether favorable or unfavorable, across the rest of the year. Budgets at clients stay whole when it comes to AI and other strategic priorities. Discretionary outlays unrelated to AI remain subdued. Consistent with prior years, we anticipate producing substantial free cash flows during the back half, and our free cash flow conversion rate is projected to exceed 100% in both Q3 and Q4. Because free cash generation was weaker during the first six months, our expectation for the free cash flow conversion rate in 2026 now sits at roughly 70%, which is beneath the 80%-90% conversion rate we typically see. Our guidance, as always, presumes we can keep delivering out of our Ukraine delivery centers at productivity levels comparable to what was achieved in 2025. Turning to the outlook for the whole year. Revenue growth is now set to fall in the range of 3.2%-4.2%. Foreign exchange should contribute a positive impact of roughly 1.2%. Organic constant currency growth is now projected to fall in the range of 2%-3%. GAAP income from operations is now anticipated to fall in the range of 10.5%-11%, while non-GAAP income from operations should fall in the range of 15.5%-16%. Our GAAP effective tax rate remains expected at 27%. The non-GAAP effective tax rate stays at 24%. On earnings per share, GAAP diluted EPS is now expected to fall in the range of $8.22-$8.38 for the whole year. Non-GAAP diluted EPS is now set to fall in the range of $13.08-$13.24 for the whole year, which yields year-over-year growth above 14% at the midpoint of that range. Weighted average share count is now anticipated at 52.2 million fully diluted shares outstanding. Shifting to the Q3 2026 outlook, revenue should fall in the range of $1.410 billion-$1.425 billion, which yields year-over-year growth of 1.7% at the midpoint of that range. Our guidance incorporates a negative foreign exchange impact of 0.1% for the quarter, which yields organic constant currency growth of 1.8% at the midpoint of that range. For the third quarter, GAAP income from operations should fall in the range of 11%-12%, while non-GAAP income from operations should fall in the range of 15.5%-16.5%. The GAAP effective tax rate should come in at approximately 25%, and the non-GAAP effective tax rate should come in at approximately 24%. On earnings per share, GAAP diluted EPS should fall in the range of $2.33-$2.41 for the quarter, while non-GAAP diluted EPS should fall in the range of $3.38-$3.46 for the quarter, which yields year-over-year growth above 11% at the midpoint of that range. For diluted shares outstanding, we anticipate a weighted average of 51.4 million. Now, here are several key assumptions underpinning our GAAP to non-GAAP measures for Q3 and the rest of the year. For stock-based compensation expense, we're looking at roughly $44 million in Q3, then $45 million in Q4. Amortization of intangibles should come in at about $17 million for each of the quarters that remain. On foreign exchange, we anticipate a loss of approximately $3 million in Q3 and a loss of $1 million in Q4. Our 2025 cost optimization program wrapped up during the Q2. Because of that, going forward for the rest of the year, severance-related expenditures will no longer be adjusted for by the company, and those expenses will show up within the company's GAAP and non-GAAP results. For the tax effect of non-GAAP adjustments, expect around $14 million in Q3 and $14 million in Q4. We anticipate a negligible tax shortfall tied to stock compensation in Q3, plus a $2 million tax shortfall in Q4. One additional assumption beyond the GAAP to non-GAAP items: interest in other income is now expected to be $1 million in Q3 and $0.5 million in Q4. And finally, my ongoing gratitude goes out to all our EPAMers for their commitment and focus on serving our clients and driving results all through 2026. Operator, please open the line for questions.
Bryan Bergin
analystHello, and thank you for the question. Perhaps I could begin by breaking down the fiscal 2026 guidance, and perhaps the demand involved here, and then explore in detail the main headwinds behind this lowered outlook. Simply looking at the business information and the high-tech vertical deceleration, it does appear to go beyond anything tied to macro factors, and I believe you acknowledged that with certain idiosyncratic pressures. I'm just wondering what you observed specifically there regarding decision-making cycles, and then concerning the North American idiosyncratic softness, could you discuss what exactly you're altering to get that growth going again? Which of those are near-term remedies as opposed to longer-term ones?
Balazs Fejes
executiveHey Bryan, how's it going? Let's dive into the tough stuff first. On the whole, the demand landscape hasn't shifted much. We're still operating in that same macro environment. What began to shift around July—and to lay this out properly, I should remind everyone that the bulk of our business comes from time and material contracts, correct? It isn't annuity-based. You've got to resell or renew those contracts pretty frequently. Around July, we began noticing clients putting AI budgets first, moving priorities away from task-based offerings such as manual testing, user experience, and as I said, from JavaScript or front-end engineering. They're moving toward modernization; that transition is taking place more quickly than the replacement work is building up, right? That leaves us with a growth gap. Clients intend to carry out ramp-ups of AI-native services, drawing mainly on savings that AI is generating. Meanwhile, they're putting money into tokens, hardware, and infrastructure. Since those savings are materializing or showing up more slowly than expected—and this is as anticipated—a gap emerges. That's what we're observing. This has hit North America quite hard; it's largely limited to our SaaS client base, which sits mostly within software and high tech. Obviously, they've got others elsewhere. The majority of SaaS clients fall in the software and high-tech sector. Within that sector, we're also noticing a certain amount of project ramp downs beyond AI areas. Those ramp downs are surpassing our growth in AI, cloud, or cybersecurity; that's how the demand environment looks to us at this moment. On to your second question.
Bryan Bergin
analystOkay. Appreciate that detail.
Balazs Fejes
executiveThe next thing we were asked about is the sales situation, correct? I believe we began discussing the structural adjustments we're implementing in our route to market back during our Q1 earnings calls. We've gotten this underway. It doesn't happen overnight. There's a process involved. What we observe from our clients is that they increasingly expect business cases and commercial proposals, instead of purely engineering-focused go-to-market motions, which used to be what took precedence. I think we already addressed this at our Investor Analyst Day. It's one component of our transformation. We're developing depth in domain expertise and consulting capability, which enables us to genuinely meet these requirements; this approach has succeeded previously, and the proof lies in the EMEA growth rate, where we made that investment earlier and carried out the transformation. We established a multi-quarter program through which we're altering our approach to market, spanning everything from marketing through to sales motions, the way we prioritize clients, and how we put proposals together. We're expanding, recruiting, and constructing a larger sales organization in North America, with particular emphasis on North America.
Bryan Bergin
analystUnderstood. All right, that tracks. Jason, regarding free cash flow, could you break down a few of the elements at play when it comes to that extra working capital drag? Did anything one-off show up in contracting? I just want to get a handle on that, along with how confident you are about free cash flow getting better in the back half, and whether anything that might stretch past this year shifts when you think about conversion.
Jason Peterson
executiveNo. Once we get beyond 2026 and into 2027 and 2028, our confidence in that 80%-90% would remain high. What we observed at the close of Q2 was that certain clients had been slated to send payments during the final two days of that quarter, but the money ultimately went out during the first two days of the following quarter, Q3. In dollar terms, that was a reasonably large sum, and it apparently needed another round of scrutiny before the payment was released. That is essentially the reason free cash flow in Q2 came in below what had been anticipated. We are observing DSO at a somewhat elevated level. My sense is that this will likely persist through the balance of the year. It won't remain where it sat in Q2, but when you line up each quarter—Q1, Q2, Q3, Q4—against its counterpart in 2025, DSO will likely run a bit higher, and that likely represents a cost. There are also some extra tax-related expenses on our side, though once again, I believe this is largely limited to 2027, and likewise to 2026. Beyond 2026, my expectation is that you'll see conversion return to the 80%-90% range.
Bryan Bergin
analystOkay. That's clear. Thank you.
Maggie Nolan
analystHello, and thanks. It seems there's also somewhat of a scheduling gap when it comes to North America. You're referring to this go-to-market transformation. Could you possibly draw some comparisons between the work done in EMEA and the efforts underway in North America, and perhaps help us grasp the duration of the rebuild phase, and when an inflection point might become visible there? Anything about some of those sizable deals and their TCV, plus the capacity to see those contributing in 2027, would be useful.
Balazs Fejes
executiveHello, Maggie. I believe we touched on this already at Investor Analyst Day — that EPAM had been heavily centered on fulfillment, deeply centered on providing engineering excellence and selling to the head of engineering or the head of products. That was particularly the case in North America, where a substantial share of our software and high-tech business is concentrated. We were largely centered on satisfying that demand rather than concentrating on and formulating business solutions, which held greater prominence across the European portfolio. Different buyers must be engaged. Different propositions must be established. Engineering excellence alone simply isn't sufficient in today's climate, since clients want to obtain ROIs or genuinely achieve business objectives. That's precisely what has been accomplished in Europe, and we put investment behind it — namely, we built highly specific domain-led go-to-market motions, combining them with greater seller capabilities, meaning numbers, processes, methods, and go-to-market motions. We're carrying that investment, which we made, into North America, scaling it and beginning to build it out; we're launching transformation programs regarding how we pursue the big deals and creating a new growth movement within EPAM, one oriented differently than before, when the emphasis was almost entirely on pure engineering excellence. It must be combined with business development excellence. As for how long this takes, I'd say it's difficult to estimate; it's in fact taking likely more time than we expected, yet we're achieving major strides this year; regarding big deals, I believe Jason provided an update, and also noted that things are advancing quite well. The count of big deals keeps growing. We're quite optimistic, and we're observing them move through the pipeline. Since our experience with certain of these matters is thinner, that's the reason we remain hesitant to fold that into our guidance, or to bring it into our forecast too early. They're moving through the pipeline, and we're determined to keep building that pipeline for the future, and this will form part of our standard go-to-market motions moving ahead. Anticipate seeing the results of this in 2027, in the form of meaningful revenue contribution.
Maggie Nolan
analystAll right, thanks. So Jason, could you perhaps speak a bit about how sustainable those margins are — things like currency movements and cost efficiencies — and can you keep margins at these levels through 2027 assuming organic growth stays down in the low single digits?
Jason Peterson
executiveSure. I think I'll avoid offering any view on 2027, though I'm glad to discuss 2026 and where gross margin stands. The truth is, currency isn't doing much for us. Even though we employ a large workforce in India, that country's share of our overall costs is far less than what many rivals carry; we still bear substantial costs in countries such as Poland, Hungary, Mexico, and even Colombia, and all of those currencies have strengthened. So currency has really not helped us. In fact, it's been a bit of a drag. What has helped is some of the pricing we took early in the year. On top of that, we've been pushing to lift profitability in our fixed fee business, our non-T&M business. That has shown solid improvement versus a year ago. We keep doing all the work we've done to further boost cost efficiency in places such as India. I'm pleased with how gross margin came in during the first half. I also anticipate gross margin will stay above 32% in each of the quarters, Q3 and Q4. Once more, we're working to do the right things to protect profitability and EPS for the rest of the year.
Maggie Nolan
analystThank you.
Puneet Jain
analystHey, thanks for taking my question. I wanted to follow up on the earlier question about go-to-market. Would you characterize it as a capability gap, versus the more business use cases and outcomes that clients are seeking? Or would you characterize it as simply a matter of messaging, positioning, and account coverage in North America?
Balazs Fejes
executivePuneet, I'd say it's a capability gap in business development. It isn't a capability gap when it comes to delivery. We can in fact deliver it, since we deliver from a global workforce, and it's been proven that in Europe, with a different go-to-market motion and different business development capability, you can in fact deliver double-digit growth, which we delivered this quarter. I'm in fact very optimistic that if we fix our go-to-market motion in North America, if we grow our growth organization, if we retarget them, equip them with other tools, then the capabilities, the product which we're bringing to the market, that product resonates with our clients; we serve global organizations, so I think what works in Europe in terms of the product, what we're selling, will also deliver the same type of growth or similar growth figures in North America if we bring it to the market in a correct way.
Puneet Jain
analystUnderstood. After that, the financial services vertical you have grew by a double-digit rate and expanded very quickly compared with the other parts of the business.
Balazs Fejes
executiveThat's right.
Puneet Jain
analystWhat is behind the stronger expansion in that area, and might this vertical indicate that the remaining parts of the business could see improved growth rates as well?
Balazs Fejes
executivePuneet, in financial services we managed to bring together our sector expertise and our AI-native abilities, and in this instance we are really running big transformation initiatives; beyond that, we are updating legacy systems at our clients through AI, drawing on EPAM's intellectual property — specifically what I referred to as MF Lens. In fact, every one of these things serves as a forerunner of what we term AI foundation elements; we have simply done a better job of bundling it and delivering it to our customers, and those customers who possess substantial legacy environments are reacting well to it. Additionally, they are constructing the data platforms along with data products that are required to anchor the AI models at the point when those models are rolled out.
Puneet Jain
analystUnderstood. Many thanks.
Jonathan Lee
analystExcellent. I appreciate you fielding these. The demand shift landing in July, after the 2Q framework had already been locked in during early May, naturally makes people wonder about how far ahead you can actually see. Since the bulk of your revenue comes from T&M instead of annuity-style arrangements, and customers are able to shuffle their budget priorities fairly fast, what's the right way to gauge how dependable the updated calendar 2026 outlook really is? Why are you confident that something comparable won't unfold in August or September among a separate set of clients?
Jason Peterson
executiveNo, that's a reasonable thing to ask, Jonathan. Looking at our full-year guide, the midpoint delivers — as we mentioned during the prepared remarks — 1.8% growth year over year, and reaching that midpoint simply calls for us to stay roughly level. We keep seeing expansion, as we talked about in Europe. We keep seeing expansion especially within financial services and life sciences. Our sense is that we genuinely have a solid lineup of growers across the business, even though we'll likely keep seeing some underperformance in North America as well as in the high-tech segment of the portfolio. Once again, I have a lot of confidence in our capacity to run at that midpoint or above, and then to run with flat execution as we go from Q3 into Q4. Getting to the low end of this range would take a meaningful sequential drop from Q3 to Q4. Once again, I genuinely feel quite confident that this one has been de-risked.
Jonathan Lee
analystThat's helpful context. As a quick follow-up, how much of the North America growth shortfall comes from customers shifting to larger, more diversified rivals that are able to package AI-driven cost reductions on infrastructure or managed services together with transformation projects, essentially letting the ramp pay for itself inside one commercial agreement? I'd note that EPAM's engineering services model is more concentrated, so it may not inherently provide that same pool of savings to redeploy. If that's what's happening, what's your thinking on the commercial side of the response?
Balazs Fejes
executiveThat's a fair question; what we've observed is not demand moving over to our competitors. Instead, we see a portion of spending shifting into tokens, which we're able to provide as well, though the majority of our customers purchase those straight or redirect that spending to infrastructure or GPUs. With our bigger deal go-to-market approach, our aim is in fact to take the savings our customers will realize through AI-driven services and turn them into reinvestment in their own growth. As for the scenario you describe, I have not witnessed it so far. We are not ceding business to rivals, if that's the point you're getting at.
Jonathan Lee
analystThanks for the additional context.
Tyler DuPont
analystHello there. Good morning to you both, FP and Jason. Tyler here, filling in for Jason Kupferberg. Appreciate you fielding our questions. I'd like to begin with North America. The revised guidance contains several variables, and I'm hoping you can walk through some of what's driving them. Setting aside the fact that you described the macro backdrop as mostly the same, what portion of this new outlook stems from the go-to-market capabilities you currently have in place, and what portion comes from customers redirecting their budgets away from services entirely toward other technology spending areas such as tokens and memory, which you brought up in an earlier answer?
Balazs Fejes
executiveIn my view, these two dynamics are occurring at the same time, isn't that right? Customers are attempting to purchase various items through various channels, correct? We have a large concentration of. First off, it's both, correct? Each of those falls into a separate category, I'd say. Those customers genuinely moving their spending away from us are largely contained within the SaaS customer base, which sits mostly in our software and high tech segment; these are the customers redirecting their budgets toward tokens, toward GPUs, which we obviously do not provide. For the remaining portfolio and customers, their demand is changing, correct? I pointed to a shift from, say, manual testing, front-end capabilities, from task-based arrangements toward AI native, AI-driven solutions. In this area, our ability to drive the business is insufficient. This is exactly the area where you need better business development, sharper go-to-market ideas, or a growth community to really move things ahead and actually get those same results. We've got that ability, we've got that expertise, and we have the certified engineers. We truly know how to produce outcomes for our customers. We just have to make sure we can sell it just as effectively.
Tyler DuPont
analystAll right, that's useful. I suppose an update regarding how pricing stands at the moment would be quite useful too. A few of your competitors lately have referenced what we might term tougher pricing competition when chasing business, or difficulties achieving the net pricing gains they wanted because productivity savings keep growing. Do you observe any of that right now, or barely at all?
Jason Peterson
executiveYes. Like we've discussed, price gains showed up early in the year. We continue to witness vendor consolidation efforts happening among particular clients. Usually those happen when a customer clearly anticipates gaining some financial benefit from combining suppliers. They clearly provide potential for extra revenue too, naturally. I believe each time we've discussed these bigger contracts, though we're referring to employing agentic solutions to create a cheaper offering, what benefits us is that right now we hardly compete at all in the agentic managed services space; that's a sizable revenue chance with what we think could be suitable, robust, and healthy margins; it might appear to be a price cut if you happen to be providing managed services currently. I can see why certain rivals would describe it that way. Once more, from our standpoint, it's an added opportunity and, because of that, isn't a price decrease.
Tyler DuPont
analystGreat. I appreciate all the detail you’ve shared. Thanks again.
James Friedman
analystHere. Hi. Thank you, FP, Jason. This is the sixth quarter in a row where AI native revenue has grown double digits on a sequential basis. I’m wondering if you can give a rough sense of what share of total revenue AI native represents today, and at what point we might expect it to start pulling up the overall growth rate for the consolidated company. How are you approaching that? Thank you.
Balazs Fejes
executiveI believe we’ve already said it accounts for 11% of our business at the moment — that’s AI native. It keeps expanding; it’s a segment growing very quickly. My view is that once it passes a particular level, which we think is 25%, that’s when you’ll begin to notice it lifting EPAM’s overall growth. Keep in mind, the way we label this is quite strict. What we refer to as AI native is very narrowly defined. We aren’t counting what other companies would describe as AI-assisted revenue in that figure.
James Friedman
analystThanks, FP. I'll rejoin the line later.
Balazs Fejes
executiveThanks.
Jim Schneider
analystMorning. Appreciate you taking the question. I wanted to ask if you might expand a bit on what was just said. Perhaps I didn't catch it — regarding that $160 million in AI native revenue you mentioned for Q2, is the $600 million objective you brought up before still something you anticipate reaching?
Balazs Fejes
executiveObviously, in 2026, $600 is the goal we've stated publicly. If we surpassed it, well, that would mean we moved closer to one of the KPIs we established for ourselves, correct?
Jason Peterson
executiveYeah. I think with that, FB's way of saying absolutely.
Balazs Fejes
executiveYeah.
Jason Peterson
executiveYeah.
Jim Schneider
analystAlright, understood. That's logical. When you examine how each vertical is performing, do you anticipate that financial services will keep up those higher growth rates as we move toward 2027? Perhaps regarding the bigger deals that you see carrying over into 2027, could you say which verticals those fall into, or are they spread out fairly evenly across many verticals? Could you just give a bit of detail on what verticals make up those deals? Thanks.
Jason Peterson
executiveYes. I believe we'll avoid discussing 2027, though perhaps we can say something about the large deals. What I can tell you is that for the rest of this year, we keep witnessing robust growth in financial services. Every bit of the AI-supported modernization, including programs which use EPAM IP, keeps producing solid revenue growth for the company. We also keep expecting to see healthy growth in life sciences. I'm not sure, FP, would you like to share some of the color regarding the big deals?
Balazs Fejes
executiveRegarding that large-deal segment, a substantial number originate with regulated verticals — for us that means financial services, banking, and insurance. That accounts for a sizable share. Once conversion patterns become clear, we'll come back to you with an update; our expectation is that a portion will close, including some of the bigger opportunities tied to the financial services sector.
Jim Schneider
analystThanks.
Paul Obrecht
analystHello, and thank you. Paul Obrecht here, filling in for Darrin Peller. FP, beyond the fresh big contracts coming up in things like BPO and agentic managed services, your guidance also featured some sizable agreements in your established sectors. I'm just wondering, now that we're further into the year, how those agreements have shaped up compared with what you originally anticipated.
Balazs Fejes
executiveNone of them slipped away from us. They're moving along, working their way through the pipeline we have. A number of these agreements are in fact at earlier stages and have begun to shift over. What I find more noteworthy, I'd say, is that even when it amounts to more conventional vendor consolidation—bigger agreements, say—those are setting off additional kinds of agreements too, such as agentic BPO or, shall we say, legacy mainframe transformation; we're observing that it's advancing. We aren't dropping them. They remain in the pipeline, and I believe we'll give you an update on the following earnings call regarding how well we've done at finalizing and converting them.
Jason Peterson
executiveRight. Paul, a portion of those agreements centered on AI. You're obviously witnessing successes there, among them the ones leveraging EPAM IP. There have also been certain vendor consolidation efforts in which we've secured wins, so those fall outside agentic managed services, and advancement has occurred on that front. As we noted, it's been rather uneven, with solid expansion in Europe, as we discussed regarding financial services, life sciences, and weaker results in certain other verticals and geographies.
Paul Obrecht
analystGot it. That clears things up, and it's useful context. As a quick follow-up, FP, could you share a little more about how we get to full stack agentic engineers? I'm just wondering what share of that comes from training people already here versus bringing in people from outside. And for the internal route, how much training does someone actually need to reach that point compared with the talent we have today?
Balazs Fejes
executiveForward FDEs really sit inside our sales motion. That's the piece we're working on alongside our AI partners and the hyperscalers as well, in order to build this out. EPAM is developing it; the development happens with people from within. There's a deliberate, clear plan to scale it up, since it's tied to how we take offerings to market and how we sell solutions together with our new partners. Our expectation is to grow this internally, and we've established training programs and an academy focused on it; that represents a major piece of what we're doing in 2026. I believe, as we've kept you posted on certifications, on the work we've done with Anthropic or with Google, and on what we announced regarding OpenAI, it will be obvious that EPAM holds the biggest concentration of certified engineering and software engineering professionals anywhere in the industry, and our goal is to be the source and the answer for every partner of ours when it comes to meeting and resolving their FDE needs. Beyond merely resolving them, we intend to be the source and the supplier that partners turn to, handling their clients' problems through FDEs. That's roughly our objective, that's what we're driving toward. That's what we're targeting in 2026.
Balazs Fejes
executiveWe appreciate you being with us on this Q2 earnings call. It's obvious there's a lot on our plate, and when we hold the next earnings call, we'll bring you up to speed on how things are going as we reshape our go-to-market approach, plus the major deals—those we'll get into—and the headway we intend to make in turning EPAM into an AI-native engineering services organization. Thanks so much. Until next time.
Read the full transcript via the API
You're viewing the first half of this call. Get the complete EPAM Systems, Inc. transcript — plus 131,000+ transcripts from 7,000+ companies, speaker segments, AI summaries and full-text search — through the EarningsCalls.dev API.
Get the API View API docs →This call discussed
For developers and AI pipelines
Programmatic access to EPAM Systems, Inc. earnings transcripts and 131,000+ others is available through the
EarningsCalls.dev REST API. Plans from $39.99/month — full transcripts, speaker segments,
full-text search, and the recently-added /api/v1/transcripts/recent polling endpoint for ETL pipelines.
More tools for stock research
OptionsBell: an unusual options activity scanner that tracks options flow across all US stocks and emails you when large trades appear in EPAM Systems, Inc. or any ticker on your watchlist.
StockMarketScan: a stock screener with alerts: 24 pre-built screens refreshed every 5 minutes, watchlists and email notifications. Both offer API and MCP access and fair monthly pricing.