Datacurve, a Y Combinator graduate specializing in high-quality data for AI training, has raised $15 million in Series A funding led by Mark Goldberg at Chemistry, with participation from employees at DeepMind, Vercel, Anthropic, and OpenAI. The new round follows a $2.7 million seed investment from former Coinbase CTO Balaji Srinivasan.

Founded by Serena Ge and Charley Lee, Datacurve has developed a “bounty hunter” model that enlists skilled software engineers to produce complex datasets required for post-training and reinforcement learning environments. The company has already distributed more than $1 million in rewards through this system.

As AI development increasingly depends on domain-specific, high-fidelity data, Datacurve aims to build the infrastructure layer for post-training data collection, beginning with software engineering and expanding into other expert fields such as finance, marketing, and medicine.

Featured image: Credit: Chemistry