Funding Details
Snorkel AI, a startup focused on developing training datasets and simulated environments for AI laboratories and corporations, has successfully completed a $350 million Series E funding round, boosting its valuation to $3.5 billion.
This latest investment, spearheaded by Insight Partners and S32, nearly tripled the seven-year-old company's valuation from $1.3 billion, which was established during its $100 million Series D round 17 months prior. Previous investors such as Addition, Lightspeed, Greylock, GV, and Wells Fargo also contributed to the Series E.
Evolution of Services
Initially, Snorkel offered software for automated data labeling. However, the company transitioned last year to a 'data-as-a-service' model, delivering complete datasets to clients. This approach combines synthetic data generation through its proprietary software and models with input from human subject matter experts, moving beyond a purely human-expert marketplace.
Snorkel reports an annualized revenue run rate of $375 million, representing an eighteenfold increase over the past year. This rapid expansion is attributed to the substantial and growing demand from AI labs for premium training data.
Market Growth and Revenue Models
Other companies operating as AI data labs have also experienced significant growth. Mercor's gross annualized revenue reached $2 billion, Handshake achieved $1 billion, and Micro1 scaled to $500 million. It is worth noting that for these firms, 60% to 70% of their top-line income is typically disbursed to domain specialists, meaning their net annual revenue is considerably lower than the reported gross figures.
Snorkel states that because it provides reinforcement learning (RL) environments and complete datasets, rather than primarily human labor, payments to its human experts are categorized under cost of goods sold. This differs from other models where such payments might directly impact headline revenue figures.
Company Origins
The company began its commercial operations in 2019, building upon four years of foundational research conducted by co-founder and CEO Alex Ratner and his team at a Stanford AI laboratory.
