New York, May 8, 2026, 10:10 ET
Innodata Inc. shares leapt roughly 89% early Friday, powered by a record-setting first quarter and a boost to its 2026 growth forecast. The AI data-engineering firm now sees potential new Big Tech deals generating about $51 million in revenue for this year. INOD last changed hands at $86.31, just off an intraday high of $90.13.
The shift is drawing attention as investors look for firms translating the AI boom into tangible sales. Innodata reported first-quarter revenue of $90.1 million, up 54% year over year—fueled by stronger demand for its AI-focused data services, with growth coming from larger customer programs and fresh contracts involving increasingly complex AI workflows.
The narrative turns from a simple earnings beat to something bigger: a shift in who’s buying. CEO Jack Abuhoff said Innodata’s latest Big Tech client brought in zero revenue last year. Now, by 2026, it could be their No. 2 account. Revenue from the rest of Innodata’s Big Tech base jumped 453% for the quarter.
Innodata bumped its full-year 2026 revenue growth outlook to at least 40%, a jump from the 35% figure it issued just 10 weeks back. Quarterly revenue topped analyst projections of $76.469 million, according to Benzinga, with adjusted EPS landing at 42 cents—handily clearing the 17-cent consensus bar.
Adjusted EBITDA climbed to $25.0 million, or 28% of revenue—double the $12.7 million posted a year ago, after backing out interest, taxes, depreciation, amortization, and select items. Adjusted gross margin also widened, hitting 47%.
Abuhoff described the quarter as “record-setting,” noting that Innodata’s momentum in AI data and evaluation is now driving both scale and cash flow. As of March, Innodata reported $117.4 million across cash, equivalents, and short-term investments. Its $50 million credit line from Wells Fargo? Untouched. Innodata Investor Relations
Abuhoff told listeners the latest Big Tech projects cover every phase—pre-training, mid-training, post-training, and evaluation. Simply put, these are the data steps needed to build, adjust, and test large language models before deploying them in actual products.
Innodata is pushing to diversify away from its traditional labor-intensive offerings. President and Chief Revenue Officer Rahul Singhal said the company’s agent observability platform launched in beta this quarter, securing a $1 million deal with a hyperscaler. Another 15 companies are currently evaluating the product. Agentic systems—AI that handles complex, multi-step processes and adapts to feedback—are at the heart of this push.
Federal projects are part of the story, too. According to MarketBeat, Innodata management pointed to positive reactions on its Palantir computer vision work, said it’s already started with a top federal systems integrator, and highlighted a prime contract slot on the Missile Defense Agency’s SHIELD program.
There’s a caveat here. Innodata’s latest 10-Q highlights its vulnerability to customer concentration and uneven project flow: one client brought in roughly 56% of Q1 revenue, with a second contributing 17%. Abuhoff mentioned on the analyst call that projects can “start and stop.” SEC
The company cautioned that contracts could be ended, projected volumes might fall short, and depending on project work means revenue takes a hit if clients push back, slim down, or drop programs. That’s the backdrop for Friday’s repricing—a solid quarter, but the business remains at the mercy of shifting AI budgets.
Innodata is moving to a single operating segment, ditching its former DDS, Agility, and Synodex divisions. The shift, the company says, lines up with its growing focus on AI and integrated delivery. The spotlight now turns to the $51 million customer ramp and recent platform efforts—investors are watching to see if these can ease customer concentration without erasing the margin improvement that’s fueled the stock’s recent surge.
