The thesis

Innodata sells data engineering services that sit directly inside the AI buildout, training and preparing data for enterprise AI platforms. Revenue grew 57.8% year over year, gross margin holds at 42.7%, and net margin of 14.7% shows the growth is already profitable. Analysts rate the stock a buy, yet the conviction signal is only medium at 54 out of 100. The tension is the price: at 44.7 times earnings, the market pays for durability that the numbers have not yet proven.

Why now

AI capital expenditure keeps accelerating, with Chinese tech giants ordering $35 billion in compliant Nvidia chips and buildout spending extending across regions. Innodata posts 57.8% revenue growth at a 14.7% net margin, so the spend is converting into profit now rather than in a future promise.

What to watch

Quarterly revenue growth against the 57.8% baseline and any disclosure of customer concentration in the next two filings. Gross margin relative to 42.7% and net margin relative to 14.7% over the next one to three quarters, because margin erosion would signal pricing power loss before growth loss.

Key risks

At 44.7 times earnings, any deceleration from 57.8% growth compresses the multiple and the stock at once. Revenue concentration in a handful of large AI customers means one budget cut cuts the thesis. A 42.7% gross margin in services is defensible only while demand outruns competition, and cheaper offshore or automated alternatives erode it.

Theme exposure

Upcoming catalysts