About

We were founded in 2024 as a collaboration between clinicians and researchers across academic medical centers. Some of our earlier studies looked at the impact of AI tools on the clinical reasoning performance of doctors. These studies have been published in Nature Medicine and covered by The New York Times, helping inform the broader conversation around clinical AI.
Our work aims to shape how clinical AI is tested, trusted, and deployed — not just by building systems, but by making the methods and evaluations transparent.
Developing research prototypes like AI clinical assistants enable us to answer important questions: What does good reasoning look like for AI systems? Where do they fail? How should we measure safety, reliability, or usefulness?
Our network runs multi-site studies, including randomized controlled trials, to test how these systems perform across different workflows and settings. We release our benchmarks and methods openly, so others can build on what we learn.