Prepared remarks
Good day, everyone. My name is Olivia, and I will be your conference operator today. At this time, I would like to welcome you to EPAM's second quarter 2026 earnings release conference call. I will now turn the call over to Mike Rowshandel, Head of Investor Relations.
Good morning, everyone, and thank you for joining us today on our second quarter 2026 earnings announcement. As the operator just mentioned, I'm Mike Rowshandel, Head of Investor Relations. We hope you've had an opportunity to review our earnings release we issued earlier today. If you have not, copies are available on epam.com in the Investors section. With me on today's call are Balazs Fejes, CEO and President; and Jason Peterson, Chief Financial Officer. I would like to remind those listening that some of the comments made on today's call may contain forward-looking statements. These statements are subject to risks and uncertainties as described in the company's earnings release and SEC filings. Additionally, all references to reported results that are non-GAAP measures have been reconciled to the comparable GAAP measures and are available in our quarterly earnings materials located in the Investors section of our website. With that said, I will now turn the call over to Balazs Fejes.
Thank you, Mike, and good morning, everyone. It's a pleasure to be with you all again. Since we last spoke, I've spent a quarter with clients, partners and our own teams across global delivery centers, and one theme keeps sharpening. AI is transforming everything we do, while adding more complexity across the enterprise. The gap between AI experimentation, adoption and optimization is EPAM's opportunity. Revenue growth in the second quarter came in the high end of our outlook range with continued improvement in profitability and strong adjusted earnings per share. Our pure AI native revenues accelerated, reaching more than $160 million in the quarter. Our operating momentum remains solid, and we will be direct today about the growth gaps we are experiencing, particularly in significant parts of the North American business and what we are already doing about it. We are executing the strategy we set out at the Investor Day back in March.
And this quarter is more evidence that we are progressing in the right direction. It's been a noisy and volatile few months, both broadly and for our sector especially, and that volatility in itself is a reflection of how disruptive AI and the technology continues to be. Our results this quarter support what we have been saying: AI accelerates demand for specialized talent and new ranges of capability, and we are executing against three strategic pillars to capitalize on these new demands. Starting with the first pillar, leading in AI native software engineering, we are rebuilding how we engineer software from the inside out by taking clients beyond traditional SDLC into fully integrated agentic enterprise harnesses, real-time data modernization, cyber resilience and token cost engineering. Today, our advanced tools like AI/Run, DIAL and MF Lens are running complex use cases deployed by thousands of specially trained EPAM engineers across hundreds of client engagements.
Whenever I sat down with clients this quarter, the same question came up: how to deliver value from AI and generate a positive return on investment. My response is that this is a complex question in which AI does not offer a simple answer. The reason is that AI does not decrease complexity. It adds to it across talent, architecture, process, governance and models. Coding gets automated; engineering doesn't. The better AI gets at writing code, the more the last-mile solution engineering and successful deployments matter. Every successful AI deployment we deliver surfaces new use cases and new scope, which is exactly why the foundational work of modernization, data engineering and retiring technical debt clients have carried for years presents the biggest opportunity for us, and one that is finally addressable because of AI. And yet, we know that we are still early in this cycle. And it isn't linear: some programs are starting, some are stopping and some are converting into new ways of engaging, changing the mix and altering the market.
On the second pillar, turning EPAM into a full stack AI native organization, we continue to accelerate and expand our strategic partnerships. These partnerships, along with others, are how we build a multimodal bench that's already feeding name pipeline and many of our larger multiyear deals. This quarter, we joined the OpenAI Partner Network as an OpenAI advanced partner with a path to Elite status. Together, we are building forward deployed engineering, cyber resilience and customer experience capabilities on OpenAI's frontier models with a first-year commitment to certify more than 5,000 OpenAI consultants and train over 10,000 EPAM specialists. Beyond creating the expert force, we are extending into security, managed services and industry-specific solutions across our main geographies. With Google, we have certified more than 2,000 of our 5,000-by-Q3 target under the Gemini Certified Partner Specialist program, building our capabilities on Google's Gemini Enterprise ecosystem for building, running and governing multistep AI agents.
This is on top of our multiyear GCP relationship with over 2,000 certified Google Cloud practitioners, agentic GCP marketplace solutions and award-winning offerings. With Anthropic, we are now among the top five globally certified partners with more than 5,700 certified engineers already ahead of the 5,000-by-Q3 milestone we set out. We are building toward more than 10,000 Claude-certified architects by year-end with over half of our delivery organization already through the Anthropic Academic coursework, backed by a dedicated group of 250 forward deployed engineering black belts. Notably, the practice now extends beyond team enablement into security and into specific verticals and key accounts. And finally, on the third pillar, our go-to-market transformation, we are using our AI native structure to open new go-to-market motions with a special focus right now in North America. Over the past quarter, we have launched a structured multi-quarter commercial transformation, standardizing how we prioritize and pursue our largest accounts, building more disciplined new logo pipeline management and investing in sales capabilities and training across the organization.
This is a forward investment in commercial discipline and the back-to-fundamentals execution focus we are applying across the whole business. Now let's turn to some quick Q2 highlights. Revenue grew 4.5% year-over-year on a reported basis with organic constant currency growth of 3.4%. Four of our six verticals grew year-over-year, led by Financial Services and Life Sciences & Healthcare. Our emerging verticals and consumer goods and retail and travel businesses both contributed to growth. Software & Hi-Tech and Business Information & Media both declined in the quarter. Business Information & Media's decline was driven by the completion of several client projects. Software & Hi-Tech experienced project ramp downs concentrated in non-AI services, which outweighed the growth in AI, cloud and cybersecurity work within the same vertical. Across geographies, EMEA continued to lead our growth with strong double-digit performance, while in contrast, the Americas delivered significantly lower growth.
Now turning to the demand environment. From a macro perspective, client sentiment and budget behavior are sitting in roughly the same zone as last quarter. The environment hasn't materially improved nor worsened, and we continue to see prolonged decision making as the war in the Middle East persists. Against this background, we continue to see some real areas of strength. Let me share some specifics. EMEA continues to drive strong revenue growth, driven by an active pipeline created by proactive go-to-market sales motions that we have already implemented. Financial Services was our fastest-growing vertical again this quarter, delivering growth across both EMEA and the Americas. A key driver for growth has been AI-led deployment of use cases, including mainframe modernization using EPAM IP to reverse engineer, automate and rebuild with new forward deployed capabilities. Life Sciences & Healthcare was our second fastest-growing vertical this quarter, picking up momentum over the past two quarters, led by pharma R&D and clinical trials paired with AI, and continued momentum in medtech products and services.
In Energy, our book of business is significantly larger than it was 12 months ago, primarily driven by expanding scope of services across our existing client base as well as new logo revenues. While we historically focused on upstream, we are now expanding our book of business into midstream, downstream and data center work for this vertical. Now some of the offsetting factors. Let me be direct. North America is not growing fast enough. We now expect it to operate below our expectations in the second half. Based on where business sits today, I want to be clear that this is not a story about waiting for the macro to turn; we own it. Two things are driving conditions. First, there's a genuine shift in what North American clients are buying. Demand is moving away from task-based services like manual testing, user experience, JavaScript front-end engineering and shifting towards AI-led modernization.
This transition is happening faster than the replacement work is ramping, creating a growth gap that needs to be filled even faster. Separately, Software & Hi-Tech pulled back this quarter, primarily due to project ramp downs. And while the underlying client relationships remain solid, the timing is creating a drag in this part of the portfolio. Second, our own go-to-market in the region has not been operating at the level it needs to, and this is squarely within our control. That's exactly what the multi-quarter commercial transformation I described earlier is going to address by building forward selling momentum into subsequent quarters. But for now, we would rather set expectations, honestly, than ask you to wait on a recovery we haven't yet earned. Jason will take you through what this means for the numbers. Now turning to the new big deals pipeline. We are seeing good progress here.
These AI-led opportunities are exclusively with existing clients, not new logos, and they continue to actively move through our pipeline. All of them are AI related, specifically agentic managed services and application maintenance. To be clear, none of them are signed yet, and we are deliberately not getting ahead of ourselves in factoring them in. The potential is real, and it's one of the things we are most encouraged by this quarter. What makes these deals notable is their composition, size and multiyear structure. We are using our AI native capabilities to compete for portions of existing clients' business for which EPAM hasn't been historically positioned, thereby expanding our footprint and impact. It's complementary growth on top of our core business, and we are focused on executing it. At the same time, the natural procurement cycle runs longer than our typical deals. Based on our best view today, the likelihood of closing and ramp timing, we now expect more meaningful revenue contribution starting in the first half of 2027 versus the second half of 2026.
Now turning to AI. Our data business grew faster than the rest of the business this quarter. And that foundational demand is exactly what continues to feed our AI native pipeline, underscoring our thesis of the largely untouched backlog underneath AI: technical debt, legacy modernization and foundational data and cloud work, all of which has to happen before AI can drive value for the enterprise. And yet, while AI native revenue growth is compounding nicely, extending its run of consecutive quarters of double-digit sequential growth, now representing over 11% of our business, getting it to a more sizable share of the business is going to take some time. The constraint we see isn't our ability to deliver. Our FDE teams, our AI/Run platform and our partners can absorb considerably more than the backlog we see today. What is needed for reliable traction within complex enterprises is a motion to bring AI from enablement to business change and impact.
Compared to a year ago, our progress is real and meaningful, but the industry overall is still relatively early in the process and set against a backdrop of rapidly changing and complex industry trade wins. Let me share a few client examples to help illustrate. For a leading financial services wealth management firm, we are modernizing with AI and accelerating mainframe transformation with reverse engineering over 10 million lines of code, delivering 60% time savings in reverse engineering effort while inserting new forward deployed engineering capabilities into fundamentally new engagement modalities. For a multinational beverage manufacturing company, we built a unified data platform to consolidate the fragmented enterprise data into a trusted AI foundation, enabling scale to deliver more than 80 data and AI initiatives. Now live, the company has seen over EUR 30 million in business operational impact over the past 12 months.
For a global energy commodity company, EPAM helped migrate more than 1,000 workloads to AWS with zero downtime for users, resulting in a 40% reduction in infrastructure and operational costs and a 30% improvement in operational efficiency. The project came at a sensitive time after an acquisition when the company did extract a critical application from its legacy environment and consolidate hundreds of aging on-premises servers. Since we spoke last, we have been honored to receive several key leadership distinctions. Databricks named EPAM its 2026 Consulting and Systems Integrator AI Partner of the Year, recognizing EPAM's work helping clients across industries operationalize AI and turn fragmented data into production-grade AI applications and agents. EPAM won the 2026 Fortress Cybersecurity Award in cloud security for migrating a Swiss private bank's entire IT landscape, hundreds of applications, to Microsoft Azure while meeting strict FINMA requirements.
Gartner positioned EPAM as a specialist in its emerging market quadrant for physical AI services. IDC MarketScape named EPAM a major player in its first worldwide Life Sciences R&D Strategic Consulting Services 2026 Vendor Assessment. Forrester included EPAM among the 28 most important vendors in the customer experience strategic consulting services landscape. And finally, the Wall Street Journal named EPAM one of its best companies for the future. These recognitions continue to reflect the hard work and dedication of our global teams and our unwavering commitment to delivering tangible results and outcomes for our clients. In summary, our strong second quarter reflects real progress against our multiyear strategy and AI-specific goals. We outperformed despite the macro backdrop, made tangible headways on each of our three strategic pillars and took deliberate early steps on the areas that need it most, particularly in North America.
We remain confident in our long-term strategy to become a global leader in AI transformation services, serving as a trusted AI accelerator and a partner for our clients. Our AI native and AI foundational momentum continues to build. And over time, these revenues will continue to make up a larger share of our overall business. We see this quarter as a solid step in that direction. We are clear-eyed that the second half will be uneven, particularly in North America. But our conviction in the strategy, the team and the commercial transformation is high. The work ahead of us over the next couple of quarters is to keep converting focus into results with the same discipline that got us here. Lastly, I want to thank you all for your continued commitment, trust and support. Jason, over to you.
Thank you, Balazs, and good morning, everyone. In Q2, we delivered strong quarterly results, outperforming the high end of our guidance ranges for organic constant currency revenue growth, profitability and EPS. Revenue was $1.415 billion, delivering year-over-year growth of 4.5%. On an organic constant currency basis, revenue grew 3.4% compared to the second quarter of 2025. GAAP and non-GAAP gross margins both improved year-over-year and exceeded our expectations for the quarter. With solid revenue growth and improved year-over-year profitability, GAAP income from operations grew by more than 20% and non-GAAP income from operations grew by 14.7%. GAAP diluted EPS grew by 26.3% and non-GAAP diluted EPS grew by 22%. AI native and AI foundational revenues continue to contribute to year-over-year growth. With more than $160 million in AI native revenues in the quarter, this is the sixth consecutive quarter of sequential double-digit growth.
Moving on to our Q2 industry performance. We delivered year-over-year growth across the majority of our verticals. Financial Services delivered strong growth and was our fastest-growing vertical in the quarter, up 11.5% year-over-year, driven significantly by insurance and asset management clients across both EMEA and the Americas. We continue to generate double-digit revenue growth in the vertical with significant contribution from AI modernization programs. Consumer goods, retail and travel delivered 2.3% year-over-year growth, notably driven by retail and consumer goods. Life Sciences & Healthcare delivered solid growth and was our second fastest-growing vertical in the quarter with growth of 8% on a year-over-year basis. Revenue growth in the vertical continues to be driven primarily by clients in life sciences and medtech. Notably, year-over-year revenue growth in this vertical continues to accelerate.
Software & Hi-Tech declined 1.3% year-over-year, driven significantly by the expected ramp down of a large client program as well as the shift in priorities that Balazs mentioned earlier. Business Information & Media decreased 2.1% year-over-year, driven primarily by the completion of several client projects. Our emerging verticals delivered year-over-year growth of 4.9%, primarily driven by ongoing strength in energy and manufacturing. From a geographic perspective, Americas, our largest region, representing 57% of our Q2 revenues, grew 0.5% year-over-year with strong growth in financial services revenue offset substantially by declines in the Software & Hi-Tech and Business Information & Media verticals. EMEA, comprising 41% of our Q2 revenues, increased 10.9% year-over-year and 9.4% in constant currency, with strong growth in financial services as well as contributions from travel, consumer goods and energy.
And finally, APAC, making up 2% of our revenues, decreased 0.3% year-over-year. Lastly, in Q2, revenues from our top 20 clients grew 3.1% year-over-year, while revenues from clients outside our top 20 increased 5.2%. Now as we move down the income statement, our GAAP gross margin for the quarter was 30.4% compared to 28.8% in Q2 of last year. Non-GAAP gross margin for the quarter was 32% compared to 30.1% for the same period a year ago, demonstrating our ability to improve profitability while continuing to invest in our AI capabilities and grow our AI native revenues. GAAP SG&A was 17.3% of revenues compared to 17.1% in Q2 of last year. Non-GAAP SG&A in Q2 2026 came in at 14.5% of revenue compared to 14.1% in the same period last year. GAAP income from operations was $152 million or 10.8% of revenue compared to $126 million or 9.3% of revenue in Q2 of last year and grew by 20.4% year-over-year.
Non-GAAP income from operations was $233 million or 16.4% of revenue compared to $203 million or 15% of revenue in Q2 of the previous year and grew by 14.7% year-over-year. Our GAAP effective tax rate, which includes a higher level of tax shortfalls related to stock-based compensation, came in at 26.7%, and our non-GAAP effective tax rate was 24%. Diluted earnings per share on a GAAP basis was $1.97 compared to $1.56 in Q2 of last year, a $0.41 increase year-over-year, reflecting growth of 26.3%. Our non-GAAP diluted EPS was $3.38 compared to $2.77 in Q2 of last year, a $0.61 increase year-over-year, reflecting growth of 22%. In Q2, there were approximately 52.3 million diluted shares outstanding. Turning to our cash flow and balance sheet. Cash flow from operations for Q2 was negative $2 million compared to a positive $53 million in the same quarter of 2025. Q2 cash flow was negatively impacted by higher variable compensation payments related to 2025 performance as well as an increase in DSO in the quarter.
Free cash flow was negative $18 million compared to positive free cash flow of $43 million in the same quarter last year. Cash and cash equivalents were approximately $800 million as of the end of the quarter. At the end of Q2, DSO was 82 days and compares to 76 days for Q1 2026 and 78 days for the same quarter last year. During the quarter, we repurchased approximately 1.3 million shares, which included open market purchases of approximately 800,000 shares for $80 million and approximately 0.5 million shares from the final settlement of our accelerated share repurchase that was paid in the first quarter. To date, since the initiation of our share repurchase program, we've returned approximately $1.6 billion in cash to shareholders. Moving on to operational metrics. We ended Q2 with more than 56,650 delivery professionals, reflecting total growth of 1.5% compared to Q2 2025. Our total headcount at quarter end was more than 62,850 employees.
Utilization was 78.3% compared to 78.1% in Q2 of last year and 77% in Q1 2026. Now let's turn to guidance. Before moving to the specifics of our 2026 and Q3 outlook, I'd like to provide some thoughts to help frame our guidance. We're encouraged by our performance in the second quarter and by the continued momentum in our pure AI native revenues, keeping us on track to meet our goal of $600 million in AI native revenues in 2026. We have also been able to improve company profitability and most notably gross margin while continuing to invest in our expanding AI capabilities. However, we are now expecting a slowdown in our revenue growth rate in the second half of the year. I'll try to be clear about the underpinnings of our updated outlook. As Balazs indicated, we delivered modest growth from North America in Q2 and now expect to continue to see very slow revenue growth from this geography for the remainder of the year.
We believe our lower growth in North America is largely idiosyncratic to EPAM and something we are working to address. At the same time, our emerging pipeline of larger opportunities continues to develop, and we remain encouraged by the size of this pipeline and the progress we're making in client negotiations. However, none of these deals have been closed. As a result, we are not expecting meaningful revenues from these deals to contribute to growth until early in 2027. We now expect modest sequential growth in Q3 as well as flattish revenues as we move from Q3 to Q4. Therefore, we will be lowering our revenue guidance for 2026. At the same time, based on our solid profitability in the first half of the year and updated forecast for the second half, we now expect to operate at the high end of our previous adjusted IFO range of 15% to 16% and have updated our guidance to reflect this. Compared to 90 days ago, we see no material improvement or worsening in the broader macro environment.
And we're not assuming any change, better or worse for the remainder of the year. Client budgets remain intact for AI and other strategic priorities. Non-AI discretionary spending continues to be muted. As in past years, we expect to generate significant free cash flows in the second half of the year with our free cash flow conversion rate forecasted to be above 100% in both Q3 and Q4. However, with the lower free cash generation in the first half of the year, we are now expecting our free cash flow conversion rate in 2026 to be around 70%, below our typical 80% to 90% conversion rate. As usual, our guidance assumes we can continue to deliver from our Ukraine delivery centers at productivity levels similar to those achieved in 2025. Moving to our full year outlook. Revenue growth will now be in the range of 3.2% to 4.2%. Foreign exchange is expected to have a positive impact of approximately 1.2%.
Therefore, the organic constant currency growth is now expected to be in the range of 2% to 3%. We now expect GAAP income from operations to be in the range of 10.5% to 11%, and non-GAAP income from operations to be in the range of 15.5% to 16%. We continue to expect our GAAP effective tax rate to be 27%. Our non-GAAP effective tax rate will continue to be 24%. For earnings per share, we now expect that GAAP diluted EPS will be in the range of $8.22 to $8.38 for the full year. And non-GAAP diluted EPS will now be in the range of $13.08 to $13.24 for the full year, producing year-over-year growth of over 14% at the midpoint of the range. We now expect weighted average share count of 52.2 million fully diluted shares outstanding. Moving to our Q3 2026 outlook. We expect revenue to be in the range of $1.410 billion to $1.425 billion, producing year-over-year growth of 1.7% at the midpoint of the range.
Our guidance reflects a 0.1% negative foreign exchange impact during the quarter, producing organic constant currency growth of 1.8% at the midpoint of the range. For the third quarter, we expect GAAP income from operations to be in the range of 11% to 12% and non-GAAP income from operations to be in the range of 15.5% to 16.5%. We expect our GAAP effective tax rate to be approximately 25% and our non-GAAP effective tax rate to be approximately 24%. Earnings per share, we expect GAAP diluted EPS to be in the range of $2.33 to $2.41 for the quarter, and non-GAAP diluted EPS to be in the range of $3.38 to $3.46 for the quarter, producing year-over-year growth of over 11% at the midpoint of the range. We expect a weighted average share count of 51.4 million diluted shares outstanding. Finally, a few key assumptions that support our GAAP to non-GAAP measurements for Q3 and the remainder of the year.
Stock-based compensation expense is expected to be approximately $44 million for Q3 and $45 million for Q4. Amortization of intangibles is expected to be approximately $17 million for each of the remaining quarters. The impact of foreign exchange is expected to be an approximate $3 million loss in Q3 and a $1 million loss in Q4. We completed our 2025 cost optimization program in the second quarter. As a result, for the remainder of the year, the company will no longer adjust for severance-related expenditures, and those expenses will be recognized as part of the company's GAAP and non-GAAP results. Tax effect of non-GAAP adjustments is expected to be around $14 million for Q3 and $14 million for Q4. We expect negligible tax shortfall related to stock compensation in Q3 and $2 million tax shortfall in Q4. And one more assumption outside of our GAAP to non-GAAP items. We now expect interest and other income to be $1 million in Q3 and $0.5 million in Q4. Lastly, my continued thanks to all our EPAMers for their dedication and focus on serving our clients and driving results throughout 2026. Operator, let's open the call for questions.
Questions and answers
The first question is from Bryan Bergin at TD Cowen.
Maybe I just wanted to start with unpacking fiscal '26 guide, and maybe the demand here, and dig in on the primary headwinds in this reduced outlook. Just based on the Business Information and the Hi-Tech vertical slowdown, it does seem to be more than something macro-related, and I think you owned up to that with some of the idiosyncratic pressures. Just curious what you saw there specifically in decision-making cycles, and then on the North American idiosyncratic weakness, can you talk about what you're specifically changing to restart that growth? What's near term versus kind of long-term fixes?
So let's start with the hard questions. Overall, the demand environment didn't really change, but we are still in the same macro space. What started to change somewhere in July: first of all, to really explain it, I need to remind everybody that most of our business is time-and-material contracts, right? This is not annuity-based. So you need to resell or renew the contracts quite regularly. And somewhere in July, we started to see that clients are prioritizing AI budgets, reprioritizing away from task-based services like manual testing, user experience, and as I mentioned, from JavaScript or front-end engineering. And they're shifting towards modernization. This shift is happening faster than the replacement work is ramping, creating a growth gap for us. And the clients are planning to do the ramp-ups of the AI-native services, mainly from savings from what AI is delivering. At the same time, they are funding tokens, hardware, and infrastructure.
So as the savings are realizing or appearing slower than expected, this creates a gap. So this is what we are seeing. This is actually very much impacted in North America. And it's mostly confined in our SaaS client base, which is largely in Software & Hi-Tech. In this sector also, we are seeing some level of project ramp downs outside of AI areas. And these ramp downs are outweighing our growth in AI, cloud or cybersecurity. So that's kind of the demand environment as we see it right now. The second question was what are we going to do about the sales? I think already in the Q1 earnings calls, we started to talk about the structural changes which we are making in how we go to market. And we kicked this off. It takes time. It's a process. But what we're seeing is that clients increasingly want to see business cases and commercial proposals rather than just engineering go-to-market motions, which was the priority in the past.
I think we already talked about this in our Investor Analyst Day. This is a part of our transformation. We are building depth in domain and consulting capability, which allows us to actually respond to these demands. This has worked before and the evidence is in the EMEA growth rate where we actually made this investment in the past and made the transformation. We put a multi-quarter program in place where we're changing how we go to market, from marketing all the way to sales motions, how we prioritize clients, how we create proposals. And we are building out, hiring and building a bigger sales force in North America, especially focused in North America.
Got it. Okay. Makes sense. And then, Jason, on free cash flow, can you just unpack some of the moving pieces there as far as the added working capital headwind? And was there anything kind of one-time in contracting? I want to just understand that and your confidence in the second half free cash flow improvement and whether anything maybe beyond this year changes as you think about conversion?
No. We would still have high confidence in the 80% to 90% conversion as we move past 2026 into 2027 and 2028. What we did see at the end of Q2 was we had clients who were expected to make payments in the last two days of the quarter, and those payments ended up being made in the first two days of the next quarter in Q3. So effectively, it was a fairly significant amount of money that required an additional level of review before they made payment. So that was really what drove the less-than-expected free cash flow in Q2. We are seeing a somewhat higher level of DSO. I do think that probably is going to continue throughout the year. It's not going to stay at the level it was at Q2, but I think as you compare each quarter—Q1, Q2, Q3, Q4—to the corresponding quarter in 2025, DSO is probably going to be a bit higher, and that's probably a headwind. We also have some additional tax-related expenses, but again, I think it's very much confined to 2026, and beyond 2026, you'll see a return to the 80% to 90% conversion range.
The next question is from Maggie Nolan at William Blair.
So it sounds like there's also a bit of a timing gap in North America. And so I'm wondering if you're talking about this go-to-market transformation, can you draw some parallels between what you've done in EMEA and what you're doing in North America, and maybe help us understand how long the rebuild phase is, when you might see an inflection point in North America? Any information on some of those large deals and TCV there and ability to see contribution from those in '27 would be helpful.
I think we already started to talk about this during Investor Analyst Day: EPAM was very focused on fulfillment, very much focused on delivering engineering excellence and selling to the head of engineering or head of product. And it was especially true in North America, where we have a large concentration of our Software & Hi-Tech business. We were focused on fulfilling that demand instead of focusing and formulating business solutions, which was more predominant in the European portfolio. You have to address different buyers. You have to have different propositions in place. Engineering excellence is just not enough in the current environment, as clients are looking to receive ROIs and actually realize business goals. So this is what we've done in Europe and invested in: we created specific domain-led go-to-market motions, pairing it with larger seller capabilities in terms of numbers, processes and methods.
We are bringing this investment into North America, expanding and starting to build it out. We started transformation programs on how we approach the big deals and are building out a new growth movement inside EPAM, which is focused differently than in the past, which was very much focused on purely engineering excellence. We need to pair it with business development excellence. How much time it will take? It's hard to say. It's actually taking probably longer than we anticipated, but we are making big strides during this year. In terms of big deals, as Jason updated, I also updated that it is progressing very well. Actually the number of big deals is increasing. We are very optimistic about it, and we're seeing that they're moving through the pipeline, but as we have less experience in some of these things, that's why we are still reluctant to bake them into our guidance early on. They are progressing through the pipeline and we are emphasizing continuing to build this pipeline for the future, and this will be part of our normal go-to-market motions going forward. Expect to see results of it in 2027 in terms of significant revenue contribution.
Okay, thank you. And then Jason, maybe can you comment a little bit on the margin durability, maybe what is foreign currency and cost optimization, and whether or not you can sort of hold these margin levels into 2027 if the organic growth is in sort of the low single digits range here?
I'm probably going to stay away from commenting on 2027, but happy to talk about 2026 and gross margin levels. We're actually not getting a lot of benefit from foreign exchange. Despite the fact that we've got a significant India workforce, India as a percentage of our total cost is not as significant as it is for many of our competitors. We still have very significant costs in places like Poland and Hungary and Mexico and even Colombia, where those currencies have all appreciated. So foreign exchange has not been a contributor; it has been somewhat negative. What we did get is price increases that we implemented at the beginning of the year, which has been helpful. We also have been working to improve our fixed-fee profitability and non-T&M profitability. That's improved nicely on a year-over-year basis. And then we continue to do the work to improve cost efficiency in places like India. I feel good about the gross margin performance in the first half and expect to continue to see gross margin in excess of 32% for each of Q3 and Q4. We're trying to do the right things to preserve profitability and EPS for the remainder of the year.
The next question is from Puneet Jain at JPMorgan.
I wanted to follow up on the earlier question on go-to-market. Would you say it's a capability gap versus providing more business use cases and outcomes that clients are looking for? Or would you say it's merely an issue around messaging, positioning and account coverage in North America?
I think it's a capability gap in business development, Puneet, not a delivery capability gap. We can actually deliver it because we are delivering from a global workforce, and it's proven in Europe that with a different go-to-market motion and different business development capability you can deliver double-digit growth, which we delivered this quarter. If we fix our go-to-market motion in North America, if we increase our growth organization, retarget and equip them with other tools, then the product we're bringing to market will resonate. We serve global organizations, so what works in Europe in terms of the product will also deliver similar growth in North America if we bring it to market correctly.
Got it. Then your financial services vertical was up double digits, growing at a rapid clip relative to the rest of the business. What's driving higher growth there and could that vertical be a precursor to better growth rates in the rest of the business?
We were successful in financial services by combining our domain knowledge with our AI-native capabilities and driving large transformation programs. We are modernizing using AI and utilizing EPAM IP—what I called out as MF Lens—on clients' legacy systems. All of these are elements of what we call AI foundations. We've been more successful packaging it and bringing it to clients who have large legacy estates; they are responding. They are also building out data platforms and data products, which ground AI models when implemented.
The next question is from Jonathan Lee at Guggenheim Partners.
The fact that the demand shift happened in July after you'd set the Q2 framework in early May raises questions about forward visibility. Given most of your business is T&M rather than annuity-based, and clients can reprioritize budgets relatively quickly, how should we think about the durability of the revised calendar '26 outlook? And what gives you confidence that a similar dynamic doesn't play out in August or September within a different client cohort?
It's a fair question. When we look at the full-year guide, the midpoint gives about 1.8% year-over-year growth, which requires that we remain roughly flat in the second half. We continue to see growth in Europe, particularly in financial services and life sciences. We think those are strong growth areas while we expect some ongoing underperformance in North America and the Hi-Tech portion of the portfolio. I feel confident that the midpoint is de-risked. You would need a material sequential decline from Q3 to Q4 to end up at the low end of the range, so I feel quite confident in our view.
I appreciate that color. Just as a follow-up, to what extent is the growth gap in North America a function of clients gravitating toward more diversified peers who can bundle AI-led savings on infrastructure or managed services with the transformation work, effectively self-funding the ramp within a single commercial construct? EPAM's more concentrated engineering services model doesn't necessarily offer the same savings pool to redeploy. If that's the case, how are you thinking about the commercial response?
Good question. We haven't seen demand migrate to peers in a way that shows we're losing deals directly to competitors. What we are seeing is that some spend is migrating to tokens and infrastructure—GPUs and the like—which clients are funding. With our larger-deal go-to-market approach, we want to transform the savings clients realize with AI-driven services into reinvestment for growth. To be clear, I haven't seen a broad shift of clients away to competitors; we're not losing deals for that reason.
The next question is from Tyler DuPont at Wells Fargo.
This is Tyler on for Jason Kupferberg. I want to start within North America. There are a few moving pieces in the updated outlook, so just want to ask about some of the drivers there. You mentioned macro is largely unchanged, so if we just put that to one side, how much of this updated outlook is due to your current go-to-market capability set versus clients actually shifting spend away from services altogether towards other tech priorities like tokens and hardware?
I think it's both happening at the same time. The clients who are shifting spend away from us are predominantly in our SaaS client portfolio, largely inside Software & Hi-Tech; those clients are shifting spend toward tokens and GPUs, which we don't deliver directly. The rest of the portfolio is shifting demand from manual testing and front-end work towards AI-native solutions. There our capability around business development and go-to-market is lacking. That's where we need to strengthen the sales and go-to-market motions. We have the certified engineers and the delivery capability; we need to be able to sell and commercialize it effectively.
Okay. That's helpful. And then, an update on the pricing environment would be helpful. Some peers have mentioned more competitive pricing dynamics to win work or challenges getting as much net pricing realization as they'd hoped due to increasing productivity savings. Are you seeing any of that now or not so much?
We did see price improvement at the beginning of the year. We're seeing ongoing vendor consolidation exercises across certain customers, which often expect economic advantage from consolidation. Those can give upside revenue potential as well. Regarding agentic solutions, they can produce a more cost-effective offering. For incumbents in managed services, that may look like a price compression, but for EPAM, it's an incremental opportunity because we don't currently participate heavily in agentic managed services. We view it as a significant revenue opportunity with appropriate profitability rather than a reduction in price for our core business.
The next question is from Jamie Friedman at Susquehanna.
Sixth consecutive quarter of double-digit sequential AI native revenue growth. Wondering if you could roughly size what percentage of revenue is AI native today, and when can we expect it to lift total consolidated company growth? How are you thinking about that?
We mentioned earlier that AI native is about 11% of our business today and continues to grow rapidly. It's a very fast-growing segment. I think once it crosses a certain threshold—around 5% incremental contribution over current levels—you'll start to see it lift consolidated EPAM growth materially. Remember, this is a narrow definition of AI native; we're not including broader AI-assisted revenue in that number.
The next question is from Jim Schneider at Goldman Sachs.
I was wondering if you could maybe just follow up on the last response. With the $160 million of AI native revenue you reported in Q2, do you still expect to hit the $600 million target you talked about earlier?
So clearly that's $600 million is our declared goal in 2026. If we overachieve it, then we actually went toward one of the KPIs we set ourselves.
Yes. FB's way of saying absolutely.
Okay, fine. That makes sense. As you look at the vertical performance, would you expect financial services to maintain the stronger growth rates heading into 2027? And maybe with respect to the larger deals you see pushing into '27, if you could address where those verticals are, or is it broadly dispersed across a bunch of verticals? Maybe just talk a little bit about the vertical composition of those deals.
I'm going to stay away from commenting on 2027 specifically, although I can say we'll continue to see strong growth in financial services throughout the remainder of this year. AI-supported modernization, including programs that use EPAM IP, continue to deliver solid revenue growth. We're also expecting to see good growth in life sciences.
In the big deals bucket, we have quite a few from regulated industries—financial services, banking and insurance—so a large portion is coming from that vertical. We'll update you on conversion progress in future calls.
The next question is from Paul Obrecht at Wolfe Research.
This is Paul Obrecht on for Darrin Peller. Outside of the new large deals emerging in areas such as BPO and agentic managed services, your guide also included some larger deals in your traditional markets. Just curious, as the year has progressed, how these deals have evolved relative to your initial expectations.
We haven't lost them; they are progressing in our pipeline. Some of these deals are in smaller phases and are starting to convert. More interestingly, traditional vendor consolidation deals can also trigger other types of engagements, such as agentic BPO or legacy mainframe transformation. We're seeing that progression. We will update you on our next earnings call on how successful we've been in closing and converting them.
Yes. Some deals were AI-specific and you've seen wins that use EPAM IP. We also have had wins in vendor consolidation exercises. Progress has been somewhat uneven, with good growth in Europe driven by financial services, life sciences, and less so in other verticals and geographies.
Right. That makes sense. That's helpful color. Just as a follow-up, Balazs, can you provide a bit more detail on the path forward to develop full stack agentic engineers? Just curious how much of that would be coming from internal training versus external hiring. And for the internal, what level of training is really required to get there relative to current talent?
FDEs are part of our sales motion. We're partnering with our AI partners and hyperscalers to build this capability. We're building it internally with deliberate ramp-up and training programs and academies. This is a big part of our effort in 2026. As we announced, certifications with Anthropic, Google and OpenAI are central to this effort. You're going to see EPAM aim to have one of the largest concentrations of certified software engineering professionals in the industry, serving as a source and supplier for partners and clients to solve their needs around FDEs.
There are no more questions at this time. I will now turn the call back over to Balazs Fejes for closing remarks.
Thank you very much for joining us on our Q2 earnings call. Clearly, we have a work cut out for us. On the next earnings call, we'll update you on how we're progressing in transforming our go-to-market motion, around the big deals we discussed, and the progress we make in transforming EPAM into an AI native engineering services organization. Thank you very much. See you next time.