Grid Dynamics is a Fortune 1000 technology partner that is rebranding from a broad digital-transformation house into an artificial intelligence firm. That rebrand now shows up in the revenue mix. AI work makes up 30.7 percent of total revenue for the second quarter, and the growth is led by the company's largest technology customers.
The mechanism behind the print is margin, not just volume. Revenue of $108.2 million grew 7.0 percent year over year, slightly above the high end of guidance. The adjusted profit measure the company reports, EBITDA, rose 16 percent to $14.7 million. Gross margin expanded to 36.6 percent, and management is on a stated path to add 300 basis points. This quarter is the first real evidence that the AI-native delivery model is producing that margin.
The tension is that the underlying business is still a services firm with thin operating leverage. GAAP net income is down sharply year over year, and operating cash flow slowed to $14.5 million in the first half. The forward question is whether the AI revenue mix can carry gross margin up fast enough to outpace the headcount the company has to keep hiring. The third quarter revenue guide of $112.0 to $114.0 million is the first test of that path, and the next few quarters resolve it.
Grid Dynamics is a product-centric engineering company that sells enterprise software development, data, cloud and AI delivery to large, well-capitalized customers. The relevant peer set is a mix of large-cap IT services firms such as Accenture, Cognizant and Infosys, a group of mid-cap digital product and AI engineering boutiques, and a handful of smaller AI-transformation specialists. Grid Dynamics sits at the high end of the boutique segment on scale, with roughly $412 million of revenue for the full year 2025, but well below the mega-cap consultancies on both size and diversification. That positioning matters, because the company is asking to be valued as an AI infrastructure company rather than as a traditional services firm, and the multiple it earns depends on how fast the mix shift into AI work happens.
The strategic repositioning is the load-bearing change. Management has framed AI as the core of the business, built around what it calls the GAIN engagement model, an AI-native delivery framework that rethinks team composition and engineering workflows for the AI era. The company's language is explicit that it is moving from effort-based development to AI and human collaboration optimized for global, enterprise-scale delivery. That is a deliberate attempt to lift the economics of the business, because the traditional services model is a low-margin, headcount-driven business where revenue and labor cost move together and the gross margin stays in the low-to-mid 30 percent range.
The customer mix is shifting in the same direction. Technology, Media and Telecom is now the largest vertical at 31.8 percent of second-quarter revenue, up from a year-ago share of 24.9 percent. The vertical grew 36.4 percent year over year on strong demand from the company's largest technology customers. Finance is the third vertical and is a relative laggard, down 2.6 percent as certain insurance engagements completed. The concentration in technology and financial services is both the strength and the risk, because those are the sectors where AI adoption is moving fastest, and they are also the sectors where the company's top accounts drive a large share of growth.
The company's footprint is global, with offices across the Americas, Europe and India, and a substantial share of its cost base sits in lower-cost geographies. That geographic arbitrage is a real part of the margin story, and it is also a source of structural risk, because the company has been running a multi-year exit from certain Eastern European operations in the wake of the Russia-Ukraine conflict. The ongoing geographic reorganization and restructuring charges are the cost of that reshaping, and they are a recurring drag on the GAAP numbers even as the underlying delivery margins improve.
The company's product story is organized around the GAIN platform family, a set of AI-based delivery tools that management says are winning wider enterprise adoption as clients move AI workloads from pilots into production. The GAIN model is not a single software product so much as a delivery methodology and a suite of supporting tooling, and it is the vehicle through which the company expects to deliver more output per unit of labor cost. That is the core of the margin thesis, because if AI tools let the same headcount deliver more revenue, the gross margin expands even when revenue and labor cost both grow.
The second-quarter print is the first where that thesis is visible in the numbers. AI revenue crossed 30 percent of total company revenue for the first time and grew more than 50 percent year over year for a second consecutive quarter, which management says is the result of clients transitioning enterprise AI workloads from pilots to production. The top accounts continue to drive growth, with several of them delivering double-digit quarter-over-quarter growth. The strategic implication is that the company is not just selling AI projects as one-off services engagements, it is embedding a repeatable AI delivery model that should stick with the customer and scale with them.
The Ekumen acquisition is the most concrete expression of the company's move into physical AI. The company closed the deal in May 2026, acquiring Ekumen, a Buenos Aires-based robotics and physical AI product engineering group, for $18.4 million. The deal gives the company end-to-end depth from robotics software through enterprise-scale deployment, and it is a deliberate bet that physical AI, the application of AI to robotics and the physical world, is the next frontier beyond the software and data work the company has been known for.
The moat is modest but real. It is built on roughly two decades of enterprise engineering experience, a productized AI delivery methodology that is harder to copy than a generic staff-augmentation offer, and a relationship base with large, sticky enterprise customers. The moat is not deep in the way a proprietary model or a licensed IP portfolio would be, and the company's own risk factors acknowledge that it operates in a rapidly evolving industry with intense competition. The durable advantage is the combination of the delivery methodology, the global cost-efficient talent base, and the top-account relationships, and the test is whether that combination produces a margin profile that the larger, less specialized services firms cannot match.
The revenue line is the cleanest part of the print. Second-quarter revenue of $108.2 million grew 7.0 percent year over year, slightly above the high end of the guidance range provided in April. The growth is concentrated in the technology and media and telecom vertical, which added $9.2 million year over year, and it is offset by a softening in the finance and other verticals. The full-year revenue guide implies roughly mid-single-digit growth over the reported full-year 2025 base, which is a modest but positive trajectory for a services business.
The margin line is where the story lives. GAAP gross profit of $39.6 million came in at 36.6 percent of revenue. That is an expansion of 250 basis points from the prior-year quarter. The company attributes the expansion to revenue growth outpacing delivery cost, improved delivery resource utilization and favorable foreign exchange. That is the first hard evidence that the AI-native delivery model is doing something, and it is the variable that determines whether the company can justify a richer multiple. The non-GAAP gross margin, which excludes stock-based compensation, was 36.9 percent, a similar improvement.
The operating expense line is the offset. General and administrative expenses rose 19.4 percent to $24.8 million, driven by higher restructuring costs, professional fees associated with potential acquisitions, and increased facility and IT costs. The result is that GAAP operating income is a thin sliver of revenue, and the company is still loss-making on a GAAP operating basis for the first half. The GAAP net income of $2.9 million is heavily supported by other income, which was $3.2 million in the quarter. That line is down from a year-ago $7.4 million on lower contingent-consideration fair-value gains and lower money-market income.
The balance sheet is the strength. Cash and cash equivalents of $298.4 million is a large buffer for a company of this size. The balance is down from $342.1 million a year earlier on the Ekumen acquisition and share repurchases. The company has no debt outstanding under its revolving credit facility, and it initiated a $50.0 million share repurchase program in the fourth quarter of 2025. It had bought back $28.8 million of that by the end of the first half. Operating cash flow of $14.5 million in the first half is positive but down from a year-ago $23.7 million, a decline the company attributes to working-capital timing. The principal financial observation is that the company is generating real cash, holding a large war chest, and returning capital to shareholders, and the question is whether the margin expansion can turn that cash generation into a durable earnings engine.
The third-quarter revenue guide implies sequential growth of roughly 3 to 5 percent. The non-GAAP EBITDA guide implies a margin near 15 to 16 percent. That is up from a year-ago 13.6 percent. That is the margin trajectory management is pointing at, and it is the number that matters most for the next six to twelve months. The full-year revenue guide is the longer-horizon commitment, and it requires the company to hold roughly 10 percent growth in the second half to land at the midpoint.
The margin commitment is the explicit strategic target. Management says it is on track to deliver a 300 basis point margin commitment, and the second-quarter gross margin expansion is the first evidence that the AI-native delivery productivity gains are showing up in the results. The risk is that the commitment is a multi-quarter process, and the operating expense line, which is still growing faster than revenue, has to bend for the commitment to be credible. The restructuring and geographic reorganization costs that have been dragging on the GAAP numbers should normalize as the Eastern European exit winds down, but that is a slow process and it is not fully under the company's control.
The Ekumen integration is a near-term execution risk that the market is not yet pricing in. The company has a maximum of $2.5 million in contingent cash consideration tied to revenue and gross-margin metrics within 12 months, which means management has to hit specific performance targets on the acquired business to unlock the full purchase price. The goodwill of $7.7 million and intangible assets of $11.0 million recognized in the deal are provisional, and the purchase price allocation is not final. If the Ekumen business underperforms, the company faces both a potential impairment and a loss of the physical-AI narrative that is part of its repositioning.
The customer-concentration risk is the structural one. The company's top accounts drive a large share of growth, and several of them delivered double-digit quarter-over-quarter growth in the second quarter. That concentration is a two-edged sword, because the same relationships that are producing the growth are also the ones that could slow down if a large technology or financial-services client deprioritizes its AI budget. The company's own risk factors note that revenue is highly dependent on a limited number of clients and industries, and that any decrease in demand for outsourced services, including from AI itself, could reduce revenue. The forward question is whether the AI revenue mix can diversify the customer base enough to reduce that concentration risk.
The first downside scenario is that the margin expansion stalls. The company's repositioning depends on the AI-native delivery model producing real productivity gains, and if those gains do not show up in the gross margin over the next two to three quarters, the margin commitment loses credibility and the multiple compresses. The data signal is the gross margin trajectory, and the test is whether it holds above 36 percent and trends toward the low 40s. A stall at 35 to 36 percent would mean the AI productivity story is not yet real, and the company is still a low-margin services business with a rebrand.
The second scenario is customer-concentration risk. If a top technology or financial-services account deprioritizes its AI spend, the revenue growth rate could fall well below the guide, and the margin trajectory could be disrupted because the large accounts are the ones with the most AI work. The data signal is the vertical mix, and the test is whether the technology and media and telecom vertical continues to grow double digits and whether the finance vertical stabilizes. A sharp deceleration in the largest vertical would be the clearest sign of the concentration risk materializing.
The third scenario is the execution risk on the Ekumen integration. The company has just acquired a robotics and physical-AI business, and the integration is not complete. The contingent consideration is tied to specific revenue and gross-margin targets within 12 months, and if those targets are missed, the company faces both a financial drag and a strategic setback. The data signal is the Ekumen revenue contribution in the next two quarters, and the test is whether the acquired business is growing and profitable enough to justify the purchase price.
The fourth scenario is the geographic, restructuring and cash-conversion tail. The company has been running a multi-year exit from certain Eastern European operations, and the geographic reorganization and restructuring charges are a recurring drag on the GAAP numbers. If the conflict environment worsens or the exit costs run higher than expected, the GAAP profitability could be depressed for longer than the underlying business warrants. Operating cash flow was also down year over year in the first half, and if the working-capital timing issues persist or the company invests more heavily in the AI delivery model, the cash buffer could erode faster than the earnings trajectory justifies. The data signals are the quarterly restructuring and geographic reorganization line items, the operating cash flow trend, and the cash balance, and the test is whether those all stabilize as the reorganization winds down.
The company trades at a market capitalization of roughly $647 million. That is based on a share price near $8 and the 81.1 million shares outstanding as of the most recent close. The value is roughly 1.6 times the full-year 2025 revenue, which is a modest multiple for an AI-focused technology company but a premium to the multiple the company earned as a traditional services firm. The valuation is being carried by the repositioning, and the question is whether the multiple is justified by the margin trajectory or by the AI narrative.
On an earnings basis, the multiple is hard to anchor. The GAAP net income for the full year 2025 was $9.7 million. That implies a GAAP earnings multiple near 67 times, a number that is not useful for a services company with large non-cash and non-operating items. The more relevant measure is the non-GAAP EBITDA, which for the trailing twelve months is roughly $58 million. That implies an EBITDA multiple near 11 times, a reasonable number for a mid-cap technology services firm and the multiple the market is effectively underwriting. The comparison set is the mid-cap digital product and AI engineering boutiques, which trade in a similar 8 to 14 times EBITDA range, and the larger consultancies, which trade at a discount because of their lower growth and lower margin profile.
The bear case is that the company is a low-margin services business with a rebrand, and the 11 times EBITDA multiple is too rich for a business whose GAAP operating margin is still near zero. In that scenario the margin expansion stalls, the customer concentration bites, and the multiple compresses toward single digits, a drawdown of 30 to 40 percent. The base case is that the AI revenue mix continues to grow at a rapid pace, the gross margin expands toward the high 30s, and the EBITDA multiple holds near 10 to 12 times, which supports the current market value. The bull case is that the AI-native delivery model produces a genuine step-change in margin, the Ekumen physical-AI business becomes a meaningful growth driver, and the company re-rates toward a double-digit multiple, a 20 to 30 percent re-rating. In short, the spread between the bear and bull outcomes is wide enough that the margin path, not the narrative, is what the market should be underwriting.
The valuation is ultimately a bet on the margin trajectory. The company is not cheap, and it is not expensive, and the difference between the three scenarios is almost entirely in the gross-margin and EBITDA-margin path over the next four to six quarters. The monitoring variables are the gross margin, the EBITDA margin, the AI revenue share of total revenue, and the customer-concentration metrics, and those are the numbers that determine whether the current multiple is a floor or a ceiling.
Grid Dynamics is a services company that is betting its multiple on an AI rebrand, and the second-quarter print is the first where the rebrand shows up in the revenue mix and the gross margin simultaneously. The company is not yet an AI company in the way a model provider or a pure-play AI software firm is, and the moat is modest, but the margin trajectory and the customer mix are moving in the right direction, and the $298.4 million cash balance gives it the runway to execute.
The central strategic initiative is the AI-native delivery model, and the test is whether it produces a durable margin expansion that justifies the repositioning. The Ekumen acquisition is the second initiative, and it is a deliberate bet on physical AI as the next frontier. The share repurchase program is the third, and it is a signal that management believes the current share price is below the value of the repositioned business. The three initiatives are interdependent, and the success of the repositioning depends on all three working together over the next four to six quarters.
The monitoring variables that the next six to twelve months resolve are the gross-margin trajectory and whether it trends toward the low 40 percent range, the AI revenue share of total revenue and whether it keeps growing at a rapid pace, the customer-concentration metrics and whether the top-account growth diversifies, and the Ekumen integration and whether the acquired business hits its contingent-consideration targets. The thesis is that the AI revenue mix is carrying the margin up, and the data that confirms or refutes it is in the next three quarterly prints.