← back to ranking

EPAM · EPAM Systems, Inc.

Information Technology Services · mkt cap $5.4B · calls: Q1 FY2026 vs Q4 FY2025
80.0 conviction · conf-adj 80

conf 6/10 partial

enthusiasm:27.0 · trend:8 · quantifies:12 · impact:4 · under_radar:14 · credibility:12 · business_impact:8 · disruption:-14 · commitment:6 · confirmation:3

Enthusiasm latest 9 / prev 9 (rising)

EPAM frames AI as a structural build-cycle opportunity—not a headcount-compression threat—via a tightly defined “pure AI native” revenue line (~$105M in Q4, >$125M in Q1, ~20% sequential, targeting $600M in 2026) plus larger “AI foundational” work. Management quantifies aggressively (certifications, project counts, client outcomes, pipeline scale) and in Q1 adds Anthropic GTM, token-aware commercial models, and ~10 outsized consolidation deals, while acknowledging budget rotation away from legacy builds and proactive “dark factory” automation pitches. Credibility is strengthened by disclosed definitions, profitability above average on AI native, and named client metrics; offset by macro-delayed discretionary spend and partial risk-adjusting of large AI deals in guidance.

GROUNDED NEXT-FY IMPACT vs CONSENSUS

Grounded on actual base — revenue $5.5B · net income $0.4B · net margin 6.9% · diluted EPS 6.72

These are next-fiscal-year annual uplift estimates, not next-quarter numbers.

Aggregate next-FY est. rev uplift: 4.64% · next-FY EPS uplift: 5.84% · vs analysts: inline · priced in: low (model's call-read: medium; verdict above is the hard-data one used for ranking) · confidence: 6/10

ClaimFigureArithmeticNext-FY Rev %Next-FY EPS %
$600M FY26 pure-AI-native revenue target
revenue		$600M FY2026 (vs ~$345-350M est. FY2025)FY25 AI base back-calced from disclosed Q4'25 >$105M + Q1'26 >$125M + 5-qtr double-digit-sequential streak = ~$345-350M; incremental FY26 = 600-~347 = ~$253M; ~253/5,457.1 = 4.64% rev; @~14.9% incr net margin (adj 11.8% + ~300bp AI premium per mgmt 'AI-native runs higher profitability') -> ~$37.6M / $642.0M adj NI = 5.84% EPS. Total AI share FY26 = 600/5,457 = 11.0%.4.645.84
Pure AI native revenue Q1'26 >$125M/qtr
revenue		>$125M/qtrComponent of the $600M annual stream; sized in the target row to avoid double-count. Annualized run-rate floor ~$500M (9.2% of rev).
Pure AI native revenue Q4'25 >$105M/qtr
revenue		>$105M/qtrHard anchor used to back-calc FY25 AI book (~$345-350M); component of $600M stream, not additive.
Sequential AI growth ~20% Q1'26 vs Q4'25
revenue		~20% QoQ125/105=1.19; growth of the same stream already captured in $600M FY26 target/phasing. Not separately additive.
5th consecutive qtr double-digit sequential AI growth
engagement · soft		5 qtrsMomentum/durability signal for $600M stream; no incremental $ beyond growth already captured.
AI-native segment profitability above EPAM average
other		> co. avg marginInforms the ~14.9% incr net margin used in the target row (vs ~11.8% adj corporate avg); margin premium is embedded in the $600M EPS calc, not separately additive to avoid double-count.
~10 large AI vendor-consolidation pipeline deals
revenue · soft		close to 10 oppsMgmt explicitly states NOT fully in outlook; no per-deal $ disclosed -> upside optionality, cannot phase to FY26.
AI-native POC-to-program conversion 60-70%
engagement · soft		60-70%Conversion rate on active projects; no pipeline ACV/timing to size FY26 revenue.
>100 new AI-native projects launched Q1
engagement · soft		>100Project count without average contract value; feeds $600M stream.
>80% of top-100 clients engaged in AI
engagement · soft		>80%Penetration breadth without top-100 revenue base or incremental ACV per client; revenue already captured in $600M.
Low single-digit rate increases (significant # of clients)
revenue · soft		low single-digitTopline lever but 'significant number' is vague — no client/revenue base to apply % to.
Fixed-fee mix 20.2% (+150bps YoY)
other · soft		20.2%, +150bpsMargin-mix shift tied to AI delivery; no disclosed margin delta to monetize.
>1,300 Claude certs (target 5,000 Q3 / 10,000 EOY)
engagement · soft		1,300 -> 10,000Delivery-capacity/capability metric; no revenue-per-cert disclosed.
>20,000 EPAMers AI-trained; 250 FDE Black Belts
engagement · soft		20,000 / 250Internal capability buildout; no productivity-to-$ conversion disclosed.
Nelnet 31% productivity / ~2x back-end speed
productivity · soft		31% / 2xCLIENT outcome, not EPAM's own cost/revenue base — proof point only.
Insurance claims processing time -75%
productivity · soft		-75%Client operational outcome, not EPAM financials — proof point.
Bayer ML pricing tool EUR20-30M incremental profit
productivity · soft		EUR20-30MThe CLIENT's (Bayer's) profit, not EPAM's — proof point, not EPAM EPS.

Assumptions: EPS sized vs consensus ADJUSTED basis ($642.0M NI / $11.42 EPS), not depressed GAAP ($378M/$6.72), per earnings-basis rule. FY2025 AI-native base (~$345-350M) reconstructed from disclosed Q4'25 >$105M + Q1'26 >$125M + the 5-qtr double-digit-sequential-growth streak. Incremental FY26 AI revenue = $600M − ~$347M = ~$253M treated as incremental to total revenue (note: some is mix-shift, and AI productivity is partly revenue-deflationary on the T&M base — a downside not netted out). Incremental net margin ~14.9% (X 15% / Y 14.76% averaged; ~300bp above the ~11.8% adjusted corporate average per mgmt 'AI-native runs higher profitability'). Margin premium embedded in the $600M target EPS, not double-counted as a separate row. Tax not separately applied (margin already net). Client-outcome metrics (Nelnet, Bayer, insurance) excluded as the client's economics. Capability/certification/engagement metrics excluded as unmonetizable. Vendor-consolidation pipeline (~10 deals) excluded as explicitly not-in-outlook and unquantified.

Top line: The only hard, sizable lever is the $600M FY26 AI-native revenue target. Against a reconstructed ~$347M FY25 base, that is ~$253M of incremental revenue (~4.6% of the $5,457M base) flowing to ~5.8% adj-EPS uplift at a modest AI margin premium. Everything else (pipeline deals, certifications, client proof points, engagement breadth) is real but unanchored optionality, not separately sizable.

Bottom line: At a 15% incremental net margin (above the 11.8% adjusted corporate average, consistent with mgmt's claim that the AI-native portfolio runs higher profitability), ~$250M incremental revenue -> ~$37.5M incremental net income = ~5.8% of the $642M adjusted net income base (~$0.67 on $11.42 adjusted EPS). Meaningful but not transformational. GAAP EPS is intentionally not used as the denominator: the depressed $378M/$6.72 GAAP base would inflate the % into a misleading figure.

Math: $600M FY26 AI target − ~$350M FY25 est. = ~$250M incremental = 4.6% of $5,457M revenue; @15% incr net margin = $37.5M = 5.8% of $642M adjusted NI ($0.67 on $11.42). Consensus shows 2024->2025 adj EPS 10.77->11.42 (+6%); a 5.8% AI-driven lift is roughly the magnitude of growth management has already guided publicly ($600M is a stated target, not a surprise), so it is largely embedded in any 2026 model — hence medium/priced-in and inline. The asymmetry sits in the ~10 vendor-consolidation deals that mgmt says are NOT fully in the outlook and the Anthropic-partnership acceleration: real upside, but unquantified, so it cannot move the sized estimate.

MODEL CONSENSUS (impact)

partial

Both agree: adopter, $600M FY26 target is sole sizable lever, ~4.6% rev / ~5.8% adj-EPS uplift. Differences are rounding on the same reconstructed base; sub-rows nulled to avoid double-count.

Conflicts reconciled
FieldOpus 4.8GPT-5.5
Rev uplift %4.6–
EPS uplift %5.8–
Priced inmedium–
vs analystsinline–
Confidence5–
Top lineThe only hard, sizable lever is the $600M FY26 AI-native revenue target. Against a reconstructed ~$350M FY25 base, that is ~$250M of incremental revenue = ~4.6% of the $5,457M base — a real but modest single-digit topline contribution, and one management already guides to ('strong line of sight'). Net topline lift is likely lower than 4.6% because AI productivity is partly revenue-deflationary on the T&M business, which EPAM is countering with fixed-fee mix (+150bps to 20.2%) and rate increases. The genuine un-modeled upside is the ~10 large AI vendor-consolidation opportunities, explicitly excluded from outlook and unquantified.–
Bottom lineAt a 15% incremental net margin (above the 11.8% adjusted corporate average, consistent with mgmt's claim that the AI-native portfolio runs higher profitability), ~$250M incremental revenue -> ~$37.5M incremental net income = ~5.8% of the $642M adjusted net income base (~$0.67 on $11.42 adjusted EPS). Meaningful but not transformational. GAAP EPS is intentionally not used as the denominator: the depressed $378M/$6.72 GAAP base would inflate the % into a misleading figure.–
ReasoningMath: $600M FY26 AI target − ~$350M FY25 est. = ~$250M incremental = 4.6% of $5,457M revenue; @15% incr net margin = $37.5M = 5.8% of $642M adjusted NI ($0.67 on $11.42). Consensus shows 2024->2025 adj EPS 10.77->11.42 (+6%); a 5.8% AI-driven lift is roughly the magnitude of growth management has already guided publicly ($600M is a stated target, not a surprise), so it is largely embedded in any 2026 model — hence medium/priced-in and inline. The asymmetry sits in the ~10 vendor-consolidation deals that mgmt says are NOT fully in the outlook and the Anthropic-partnership acceleration: real upside, but unquantified, so it cannot move the sized estimate.–

Rows highlighted where the two models disagreed.

QUANTIFICATIONS
Pure AI native revenue (quarterly): >$125 million (Q1 FY2026, topline)
“pure AI revenues exceeded $125 million in Q1, up nearly 20% sequentially from Q4”
Pure AI native revenue sequential growth: nearly 20% (Q1 FY2026 vs Q4 FY2025, topline)
“pure AI revenues exceeded $125 million in Q1, up nearly 20% sequentially from Q4”
Pure AI native revenue (annual target): $600 million (FY2026 full year, topline)
“This momentum gives us a strong line of sight to our $600 million target for the full year”
Pure AI native revenue sequential growth streak: fifth consecutive quarter of sequential double-digit growth (through Q1 FY2026, topline)
“This is the fifth consecutive quarter of sequential double-digit growth”
Claude certifications (current): >1,300 (Q1 FY2026, both)
“more than 1,300 are already Claude certified”
Claude certifications (target): 5,000 by end of Q3; 10,000 by year-end (FY2026, both)
“We expect to reach 5,000 certifications by end of Q3 with 10,000 by year-end”
Anthropic/AI training completions: >20,000 EPAMers (to date in Q1 call, both)
“over 20,000 EPAMers have completed training via entropic economy”
Forward deployed engineering Black Belts: 250 (current, both)
“specialized cadre of 250 forward deployed engineering Black Belts”
Top-100 client AI engagement: >80% (Q1 FY2026, topline)
“more than 80% of our engaged in AI initiatives”
New AI-native projects launched: >100 (Q1 FY2026, topline)
“more than 100 new AI-native project launched in Q1”
Client productivity (Nelnet / AI-Run): 31% productivity increase; ~2x back-end development speed (client program outcome, both)
“Nelnet achieved a 31% productivity increase, accelerated back-end development by nearly 2x”
Insurance claims processing time: 75% decrease (post-implementation, both)
“time to process first order of loss events decreased by 75%”
Large AI/vendor-consolidation pipeline opportunities: close to 10 (Q1 FY2026, topline)
“We're actually talking about close to 10 opportunities at this point of time”
AI native segment profitability vs. corporate average: higher than EPAM average (current (~$125M+/quarter portfolio), bottomline)
“our current AI native business or portfolio, which is over $125 million per quarter is run higher profitability than EPAM average”
Pure AI native revenue (quarterly): >$105 million (Q4 FY2025, topline)
“in Q4, we generated more than $105 million in pure AI native revenues”
Pure AI native revenue (annual target): in excess of $600 million (FY2026, topline)
“expect to scale these revenues in excess of $600 million in 2026”
AI native POC-to-program conversion: 60% to 70% (Q4 FY2025 active projects, topline)
“between 60% to 70% have expanded from initial proof of concept into larger programs”
Client ML pricing tool profit uplift (Bayer): EUR 20 million to EUR 30 million incremental yearly profit (annual client outcome, both)
“delivered EUR 20 million to EUR 30 million in incremental yearly profit”
Fixed-fee revenue mix: 20.2% (150 bps YoY increase) (Q4 FY2025, both)
“150 basis point increase in fixed price as a percentage of total revenue to 20.2%”
Pricing — clients receiving increases: low single-digit rate increases (significant number of clients) (entering 2026, both)
“quite significant number of clients in both Europe and North America are giving us at least low single-digit rate increases”
PAST (realized)
CURRENT (now)
FORWARD (guidance)
TRACK RECORD — PROMISE vs DELIVERY

72/100 track record   too-early  6 calls reviewed

Across six calls EPAM rarely issues dated numeric AI targets; the main forward commitment is >$600M pure AI-native revenue in 2026, which looks early-on-track ($125M in Q1) but is not yet provable. Reporting metrics (AI-native revenue scale, client engagement, POC-to-program conversion) have been consistent without obvious walk-backs, but most judgeable delivery is still pending.

>$600M pure AI-native revenue in 2026 — promised Q4 FY2025
too-early Q1 FY2026 reported >$125M pure AI-native revenue (~20% sequential growth) and management said it has strong line of sight to the $600M full-year target; FY2026 not complete.
5,000 Claude certifications by end of Q3 FY2026 — promised Q1 FY2026
too-early Management cited >1,300 Claude-certified staff in Q1 and a path to 5,000 by end-Q3; deadline not yet reached in this transcript set.
10,000 Claude certifications by year-end FY2026 — promised Q1 FY2026
too-early Same Q1 update (~1,300 certified, >20,000 Anthropic-economy trained); year-end target not yet testable.
~$50M quarterly AI-native program revenue (vs single-digit millions in H1 2024) — promised Q4 FY2024
delivered Later redefined and scaled the metric (>$105M pure AI-native in Q4 FY2025; >$125M in Q1 FY2026) with five straight quarters of double-digit sequential growth cited—not a missed dated target.
75% of top-country clients engaged on GenAI initiatives — promised Q4 FY2024
delivered Later calls kept broad top-client AI engagement (e.g., vast majority / >80% of top 100) without walking back the adoption narrative.
60%–70% of active AI-native projects expanded from POC to larger programs — promised Q3 FY2025
delivered Same 60%–70% range repeated in Q4 FY2025 and Q1 FY2026 as an ongoing operating metric, not abandoned.
PRICED-IN (REFINED)
LOW (room left)

Est. revisions flat  ·  Fwd P/E 14.1  ·  EV/Sales 0.8x

AI claim maps to Software And Hi-Tech Sector, Financial Services Sector, Healthcare Sector

Analyst ratings are broadly stable (buy-heavy, with a slight June shift toward more holds) while price targets are flat month-over-quarter at $137 but well below the $166 last-year average, so revision momentum is not rising. Forward P/E ~14x and EV/Sales ~0.8x are modest for IT services, indicating the market is not paying a premium for AI-driven upside. AI-related demand and delivery efficiency would most plausibly flow through Software And Hi-Tech and large verticals (Financial Services, Healthcare). Flat-to-soft revisions plus depressed multiples imply AI upside is not yet fully priced in.
COVERAGE — ENTHUSIASM TRAJECTORY + CATALYSTS
8Q4 FY20247Q1 FY20258Q2 FY20259Q3 FY202510Q4 FY20259Q1 FY2026

AI enthusiasm across 6 calls — trend ↗ rising

Shifted from client GenAI traction to quantified pure-AI revenue, $600M target, literacy metrics, and named platforms/partners.

RECENT AI CATALYSTS & NEWS
BUSINESS IMPACT - QUALITATIVE MATERIALITY

7/10 qualitative impact   material  near-term · hard evidence

Where AI matters: pure AI-native services revenue and delivery productivity

Disclosed pure AI-native revenue is scaling fast (>$125M/qtr, $600M FY26 target ~11% of sales) with above-average margins and real client outcomes, but incremental math is only ~4-5% revenue and ~6% adjusted EPS uplift—not a company-wide economic reset.

Caveats: $600M AI-native target may embed reclassification/mix-shift more than net-new demand; Legacy build/maintenance book faces ongoing hour deflation faster than AI-native line scales; ~10 vendor-consolidation deals are upside optionality but largely risk-adjusted out of guidance; Macro discretionary delays can stall AI programs despite strong engagement metrics

AI DISRUPTION / CANNIBALIZATION RISK  headwind · 7/10

EPAM still monetizes billable engineering hours across a large legacy T&M base, and GenAI/client 'dark factory' productivity directly compresses those units and pushes token/non-T&M pricing even as management rotates spend toward AI-native work.

OPTIONS / MARKET STRUCTURE

option liquidity: fair

ATM IV
—
TYPICAL BID-ASK
—
OPEN INTEREST
—

proxy inputs — dollar-ADV $176M · beta 1.453 · px $99.88

source: proxy (no options chain on FMP)
FMP /stable/ exposes no options-chain endpoint on this key, so ATM IV, bid-ask spread and open interest are unavailable. Liquidity below is a PROXY from dollar-ADV, beta and price level (a stand-in for option depth), not measured option-market data.

CONFIRMATION — INSIDERS · 13F · LANGUAGE
Mixed — insiders selling, institutions adding, management language 8/10 committed.
INSIDERS selling 1 open-market sell(s) vs 6 buy(s) — net distribution
INSTITUTIONS (13F) adding as of 2026-03-31: 80 new / 149 closed positions; 283 increased / 179 reduced; institutional ownership +6.41pp; -67 net 13F holders
MGMT LANGUAGE 8/10 committed Firm AI revenue metrics, certification targets, and project counts; hedges mainly on pipeline timing and scaling motions.
commit “pure AI revenues exceeded $125 million in Q1, up nearly 20% sequentially from Q4”
commit “We expect to reach 5,000 certifications by end of Q3 with 10,000 by year-end.”
commit “We had more than 100 new AI-native project launched in Q1”
VERBATIM AI QUOTES
“pure AI revenues exceeded $125 million in Q1, up nearly 20% sequentially from Q4. This momentum gives us a strong line of sight to our $600 million target for the full year”
— Balazs Fejes, Q1 FY2026
“we just announced a strategic multiyear applied AI partnership with Ontic to accelerate the delivery of safe, reliable enterprise-grade AI for our clients. As an Anthropic services partner, EPAM is building a dedicated practice for more than 10,000 cloud-certified architects, including the specialized cadre of 250 forward deployed engineering Black Belts”
— Balazs Fejes, Q1 FY2026
“To date, over 20,000 EPAMers have completed training via entropic economy and more than 1,300 are already Claude certified. We expect to reach 5,000 certifications by end of Q3 with 10,000 by year-end”
— Balazs Fejes, Q1 FY2026
“Our aspiration is to become the go-to partner for enterprise a transformation with a focus on 3 strategic pillars, which are helping reshape the company. These pillars include establishing ourselves as a leading AI delivery software engineering services provider, transforming ourselves into an AI-native organization and capitalizing on our AIT structure to expand go-to-market offerings”
— Balazs Fejes, Q1 FY2026
“With our Client Zero mentality, we are engineering an entirely new operating model one that dynamically blends human talent, AI capabilities and advanced agentic systems to run the business foster better and lower cost across all geographies”
— Balazs Fejes, Q1 FY2026
“continued shift in spend towards AI native and strategic deployments. Clients continue to turn to EPAM for help in addressing the widening adoption gap”
— Balazs Fejes, Q1 FY2026
“Importantly, our client pipeline of AI programs and fundings remain strong”
— Balazs Fejes, Q1 FY2026
“Further token usage and the associated economics are only becoming a more integral part of the investment thesis and business case”
— Balazs Fejes, Q1 FY2026
“Looking across our top 100 clients, traction remains strong as more than 80% of our engaged in AI initiatives”
— Balazs Fejes, Q1 FY2026
“we had more than 100 new AI-native project launched in Q1”
— Balazs Fejes, Q1 FY2026
“EPAM is seeing an oxalating large-deal pipeline focused on AI-enabled vendor consolidations, where EPAM has significant opportunity to gain market share”
— Balazs Fejes, Q1 FY2026
“Across our pure AI native revenues, our momentum continues and fundamentals remain intact with another quarter of double-digit sequential growth”
— Balazs Fejes, Q1 FY2026
“Nelnet achieved a 31% productivity increase, accelerated back-end development by nearly 2x”
— Balazs Fejes, Q1 FY2026
“time to process first order of loss events decreased by 75%”
— Balazs Fejes, Q1 FY2026
“AI native and AI foundational revenues continued to contribute to year-over-year growth with more than $125 million AI native revenues in the quarter. This is the fifth consecutive quarter of sequential double-digit growth”
— Jason Peterson, Q1 FY2026
“I already became available due to our AI native and AI/Run capabilities, which opened us for large vendor consolidation, large transformation deals, which is for us was outside of our normal norm”
— Jason Peterson, Q1 FY2026
“I think Anthropic is going to be a very important relationship for -- we are -- I think we are following a playbook, which we've done before”
— Balazs Fejes, Q1 FY2026
“We will be going to the market together with Anthropic and bring to the market applied AI solutions”
— Balazs Fejes, Q1 FY2026
“we are hoping to see acceleration from this partnership”
— Balazs Fejes, Q1 FY2026
“We're actually talking about close to 10 opportunities at this point of time. These opportunities are outsized in terms of range, all of them are non-T&M, so different commercial models, combining AI, token in the picture themselves”
— Balazs Fejes, Q1 FY2026
“our current AI native business or portfolio, which is over $125 million per quarter is run higher profitability than EPAM average”
— Balazs Fejes, Q1 FY2026
“in most client relationships or clients are bearing the cost of the tokens”
— Balazs Fejes, Q1 FY2026
“we are actually not seeing what we call rate compression at this point of time. We're quite successful for minority -- small minority of our clients to negotiate rate increases”
— Balazs Fejes, Q1 FY2026
“clearly, there is some impact. clients shifting some of the IT budgets towards AI spending and also the increasingly automating parts of the SDLC, for example, testing itself. And probably, they are diverting investments away from digital platform, e-commerce platform build-outs towards new AI native products or a native platforms construction”
— Balazs Fejes, Q1 FY2026
“we demonstrated dark factory capabilities. And yes, we are proactively talking to our clients, how we can introduce them how we can actually provide them a dark factory based for the autonomous applicable maintenance and support capabilities, how we can automate a large part of the testing flows”
— Balazs Fejes, Q1 FY2026
“You need to, first of all, control the model usage, what task, which model you are using what is the frequency of that model. You need to have the right blend of model”
— Balazs Fejes, Q1 FY2026
“we see a year of AI momentum marked by our clients' ongoing shift in spending towards AI investments and strategic deployments. Importantly, we expect to build on our growing momentum in AI native services, supported by our AI foundational services that enable clients to scale AI across their enterprises”
— Balazs Fejes, Q4 FY2025
“With our internal AI-native engineering transformation nearly complete, we are now shifting to develop more verticalized AI-native business offerings and consultancies”
— Balazs Fejes, Q4 FY2025
“in Q4, we generated more than $105 million in pure AI native revenues, where we continue to see solid momentum and strong sequential growth”
— Balazs Fejes, Q4 FY2025
“our AI native revenues are defined across 2 groupings, number one. AI native IP products, platforms and solutions where AI was the core of the solution versus simple work accelerated by the use of AI tools. And number two, AI-led transformation initiative across the entire enterprise. Importantly, our definition excludes all the [ IF ] foundational services along with any AI assisted work performed by EPAM employees within the software delivery life cycle”
— Balazs Fejes, Q4 FY2025
“we continue to see robust demand for our AI native services and expect to scale these revenues in excess of $600 million in 2026”
— Balazs Fejes, Q4 FY2025
“AI continues to trigger both incremental and sustained demand and is driving positives in our pipeline”
— Balazs Fejes, Q4 FY2025
“continued shift in spending towards scaled AI deployment”
— Balazs Fejes, Q4 FY2025
“between 60% to 70% have expanded from initial proof of concept into larger programs”
— Balazs Fejes, Q4 FY2025
“the size of this portfolio is already significantly larger than our pure AI native revenue base”
— Balazs Fejes, Q4 FY2025
“Leveraging machine learning, the tools delivered EUR 20 million to EUR 30 million in incremental yearly profit, reduced analytics time by [ Tenex ]”
— Balazs Fejes, Q4 FY2025
“EPAM's [ AIron ] developer agent was recently ranked in the top 5 on SWE bench verified leaderboard”
— Balazs Fejes, Q4 FY2025
“Clients need help in their AI transformation journeys and our advanced engineering capabilities, AI assets and strong delivery execution are helping clients address their most complex business challenges”
— Jason Peterson, Q4 FY2025
“strong momentum in AI native and AI foundational services”
— Jason Peterson, Q4 FY2025
“AI continues to be the net growth driver for our business”
— Balazs Fejes, Q4 FY2025
“This is going to open up a tremendous opportunity for EPAM. It's going to turn [indiscernible] the buy versus build question, right? And EPAM is a builder”
— Balazs Fejes, Q4 FY2025
“we are not seeing a pressure on our pricing due to AI. Again, most of the pricing that we have is time and materials”
— Jason Peterson, Q4 FY2025
“our definition of AI native revenue is super tight which means that we're not including a lot of things, which probably some of our competitors do include”
— Balazs Fejes, Q4 FY2025
“we expect to reach $600 million in 2026 for EPAM. So it's actually scaling up, growing very rapidly, but it's still a smaller part of our business”
— Balazs Fejes, Q4 FY2025
ANALYST QUESTIONS ON AI
Q (Q1 FY2026, Bryan Bergin (TD Cohen)): On the Anthropic relationship — how different is that model vs. heritage delivery, how difficult a pivot, and could it drive an inflection in AI native revenue growth mix?
A: Balazs Fejes: Follows prior partner playbook (engineer prep, certification ramp); over 1,400 certified cloud architects; go-to-market with Anthropic on applied AI / safe enterprise AI; "I don't think it's a pivot, it's an expansion" and "we are hoping to see acceleration from this partnership."
Q (Q1 FY2026, Jason Kupferberg (Wells Fargo)): Clarify large vendor-consolidation opportunities in 2H guide — count, risk-adjusted inclusion, and nature of work.
A: Balazs Fejes: ~10 outsized non-T&M opportunities mixing AI, tokens, business transformation and vendor consolidation; Jason Peterson: only a modest risk-adjusted subset in guide.
Q (Q1 FY2026, James Friedman (Susquehanna)): Elaborate on AI-enabled vendor consolidation mentioned in prepared remarks.
A: Balazs Fejes: Deploys AI/Run Transform on large deals; "combination of our global capabilities augmented with AI" that "challenges the status quo in the vendor landscape."
Q (Q1 FY2026, Yu Lee (Guggenheim)): Confidence closing/executing large multiyear AI deals — sales muscle, governance, margins vs. current book.
A: Balazs Fejes: Pipeline surprised management; risk-adjusted inclusion; differentiated AI-native offering; AI native portfolio "over $125 million per quarter" runs "higher profitability than EPAM average."
Q (Q1 FY2026, Bryan Keane (Citi)): How has contract pricing changed, especially Anthropic — revenue recognition vs. prior years?
A: Balazs Fejes: Token economics still evolving; most clients bear token cost; Anthropic work uses Anthropic stack; exploring commercial models; "not seeing" rate compression, with rate increases for a small minority of clients.
Q (Q1 FY2026, Arvind Ramnani (Truist)): Revenue/workflow cannibalization as AI ramps; and are you proactively offering clients headcount reductions from model improvements?
A: Balazs Fejes: Budget shift to AI, SDLC automation (e.g. testing), diversion from legacy platform builds to AI-native products; proactively pitching dark-factory / autonomous maintenance-support and automated testing flows.
Q (Q1 FY2026, James Faucette (Morgan Stanley)): On longer projects with token costs — what levers/relationships control margins?
A: Balazs Fejes: Control model usage per task, blend models for ROI, multi-source same tokens like a trading desk to manage price/availability/caps — affects "pricing and profit levels."
Q (Q4 FY2025, Margaret Nolan (William Blair)): Investments needed for vertical industry expertise — material to P&L and competitive positioning?
A: Balazs Fejes: 2026 guide already reflects investments; prioritizing BD plus industry vertical accelerators beyond AI.
Q (Q4 FY2025, David Grossman (Stifel)): When does delayed decision-making on larger AI programs break — any momentum/conversion data points?
A: Balazs Fejes: Larger programs building; some industries can't delay AI spend; procurement slows selection; will share more AI metrics; "$600 million in 2026" for AI native revenue scaling rapidly.
Q (Q4 FY2025, Bryan Keane (Citi)): AI pressure on software/IT services from Anthropic/OpenAI modules — and productivity/pricing pass-through pressure beyond pure AI revenue?
A: Balazs Fejes: Bullish — more software will be built, coding automated, EPAM wins high-end build work; Jason Peterson: No AI-driven bill-rate compression; not heavy BPO/maintenance/testing exposure.
Q (Q4 FY2025, Jim Schneider (Goldman Sachs)): Is ~80% AI native revenue run-rate growth from Q4 math directionally right, and how is EPAM's AI native revenue definition different from peers?
A: Jason Peterson: Directionally correct, strong YoY/sequential growth expected; Balazs Fejes: "Super tight" definition — only AI-core solutions and enterprise AI transformation (AI 360), excludes AI-assisted SDLC work and AI foundational/data-cloud readiness (larger than AI native).
ERROR
ask_single(composer) failed after 2 tries: [Errno 2] No such file or directory: 'cursor-agent'
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/dev/ai-earnings-scan/llm.py", line 365, in ask_json
    return ask_single(model, prompt, retries=retries, effort=effort)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/dev/ai-earnings-scan/llm.py", line 244, in ask_single
    raise RuntimeError(f"ask_single({model}) failed after {retries + 1} tries: {last}")
RuntimeError: ask_single(composer) failed after 2 tries: [Errno 2] No such file or directory: 'cursor-agent'