Datacurve raises $15 million to take on Scale AI

As the AI ​​companies are mature, the fight for high quality information has become one of the competitive areas of the industry, the marker, surge and most prominently launched by Alexandra Wang’s scale AI. But now that Wang Meta is moving forward to run AI, many funders are seeing an opening – and companies with compulsory new strategies to collect training data are willing to fund.

Wi -combinator graduated Datakurov A company that focuses on high quality data to develop software. On Thursday, the company announced a $ 15 million series led by Mark Goldberg, led by Chemistry, with the participation of Deepmind, Versal, Ethnic and OpenAI. The series comes after a $ 2.7 million seed round, which attracted investment from the former Coinbase CTO Balaji Srinivasan.

Dataccharov uses a “grace hunter” system to attract skilled software engineers to complete the source datasets. The company pays for this contribution, so far Million distributes more than $ 1 million to grants.

However, co-founder Serena GE (depicted above with co-founder Charlie Lee) has said that the biggest inspiration is not financial. For high-value services such as software development, salary for data work will always be much lower than conventional employment-the most important end of the organization is a positive user experience.

“We consider it a consumer product, not data labeling operation,” GE said. “We spend a lot of time to think about: How can we make it optimal so we want people to be interested and get on our platform?”

This is especially important because the need for the next data of training becomes more complicated. Previous models were trained in general datasets, today’s AI products depend on the complex RL environment, which needs to be built by collecting specific and strategic data. As the environment becomes more sophisticated, the data requirements are both intense for both quantities and quality-a reason that can give an edge to high quality data collection companies like Detacurov.

As an early stage agency, focusing on datacurva software engineering, but GE says that the model can easily apply to fields such as money, marketing or even drugs.

“What we are doing right now is that we are creating an infrastructure to collect post-training data that attracts and holds highly skilled people in their own domains,” GE says.