Refuel AI, a company that provides clean and labeled training data for AI models using large language models (LLMs), raised USD 5.2 million in seed funding in a round led by General Catalyst and XYZ Ventures.

The company plans to use the funding to grow its team, build out its platform, and release its product commercially by July. It also aims to invest in its open-source library and community.

Formed in 2021, Refuel AI uses LLMs to automate the creation and labeling of datasets of any volume for any business and use case. According to the company, in private beta tests, its platform sped up the data creation and labeling process by up to 100x, translating weeks or months of work down to an hour. Additionally, it has released its open-source library that enables AI teams to label their data using their preferred LLM in their own environment.