Snorkel AI valued at $3.5 billion in $350 million Series E

Snorkel AI's new round comes 17 months after it raised $100 million in a Series D at a $1.3 billion valuation, meaning the company's valuation has nearly tripled in that period. The company started commercially in 2019 after four years of research by co-founder and CEO Alex Ratner and his team at a Stanford AI lab, and it has since moved from selling data-labeling automation software to selling completed datasets and reinforcement learning environments under a data-as-a-service model.

That shift matters for how its numbers read. Snorkel says expert payments are accounted for in its cost of goods sold rather than in its headline annualized revenue, unlike data companies that pay out roughly 60% to 70% of top-line income directly to domain specialists. So the comparison with peers such as Mercor, Handshake, and Micro1 is not apples-to-apples: their headline gross figures are substantially higher than their net annual revenue, while Snorkel's headline revenue is not reduced by the same labor payouts.

The round's significance is likely to hinge on whether AI labs' appetite for high-end training data sustains Snorkel's reported growth rate, and on how investors weigh the different revenue definitions used across this emerging group of data providers. For buyers of training data, the practical question is whether a completed-dataset model delivers the quality that labs need as they build out reinforcement learning environments.