Let's Build Thoughtful AI for Healthcare

Where its needed most

April 16, 2025

Large language models like GPT-4o, Gemini, and Claude are making headlines for their potential in healthcare, powering everything from documentation tools to medical Q&A. Institutions like Stanford, Mass General Brigham, and Mayo Clinic are actively testing these models for real-world applications. While these models perform well on medical benchmarks, their plug-and-play use in clinical settings, especially in specialty care, warrants careful scrutiny.

We are builders deeply committed to solving the kinds of healthcare problems that matter, not just to the system, but to us personally. As we develop solutions at the intersection of AI, clinical reasoning, and EHR data, we’re uncovering technical challenges, risks and solutions that deserve to be shared.

This article series will chronicle those discoveries—one challenge, and one insight at a time.

Our goal is to spark conversation, collaboration, and progress across the community of clinicians, researchers, engineers, and entrepreneurs contributing to the future of healthcare. If you’re building in this space, we hope this series feels like a conversation with curious minds, who know what’s at stake and believe in getting it right.