As AI companies mature, the fight for high-quality data has become one of the most competitive areas in the industry. This has launched companies like Mercor and Surge, and most prominently, Alexandr Wang’s Scale AI. With Wang’s recent move to run AI at Meta, many investors see an opening and are now willing to fund companies with compelling new strategies for collecting training data.
The Y Combinator graduate Datacurve is one such company, focusing on high-quality data for software development. The company announced a fifteen million dollar Series A round, led by Mark Goldberg at Chemistry with participation from employees at DeepMind, Vercel, Anthropic, and OpenAI. This Series A comes after a two point seven million dollar seed round, which drew investment from former Coinbase CTO Balaji Srinivasan.
Datacurve uses a bounty hunter system to attract skilled software engineers to complete the hardest-to-source datasets. The company pays for these contributions, having distributed over one million dollars in bounties so far. However, co-founder Serena Ge says the biggest motivation for participants is not financial. For high-value services like software development, the pay for data work will always be far lower than conventional employment, so the company’s most important edge is a positive user experience.
Ge stated that they treat the platform as a consumer product, not a data labeling operation. They spend significant time thinking about how to optimize the platform so that the people they want are interested and join. This is particularly important as the needs of post-training data grow more complex. While earlier models were trained on simple datasets, today’s AI products rely on complex reinforcement learning environments which need to be constructed through specific and strategic data collection. As these environments grow more sophisticated, the data requirements become more intense for both quantity and quality, a factor that could give high-quality data collection companies like Datacurve an edge.
As an early-stage company, Datacurve is currently focused on software engineering, but Ge says the model could apply just as easily to fields like finance, marketing, or even medicine. Ge explained that what they are creating is an infrastructure for post-training data collection that attracts and retains highly competent people in their own domains.