Panacea Research was founded to turn the observations we made across our services work into training environments for frontier models.

We partner with AI labs to build long-horizon reinforcement learning environments grounded in expert regulatory judgment. We draw on our large network of ex-FDA experts to generate synthetic tasks, curate verifiable rubrics, and continuously validate them against the patterns, failure modes, and edge cases we encounter in active regulatory work.

The aim is to teach models to navigate drug and device development end-to-end.