Announcing Project ProtoStar: Making Publicly Funded Science AI-Ready
huggingface ·
Hugging Face has announced Project ProtoStar, an initiative designed to make publicly funded scientific research more accessible and reusable for AI-driven discovery. The project aims to bridge the gap between public availability and practical usability by encouraging institutions to release datasets, models, and code in AI-ready, machine-actionable formats. Founding partners include a coalition of universities such as Carnegie Mellon and Purdue, alongside industry and community organizations like ORCA and Hugging Face itself. This effort responds to federal policies that increasingly require the sharing of research data, addressing current issues where legacy formats and poor documentation hinder reuse. To demonstrate the value of this approach, ProtoStar will host a series of open challenges, starting with a Climate Risk Intelligence competition at Carnegie Mellon University on November 2, 2026. Participants in the inaugural challenge will use historical weather data to predict urban climate conditions and propose adaptation policies. The initiative seeks to build a practical evidence base for improving research infrastructure and fostering innovation across disciplines.
huggingface