Medra Launches AI Experimentalist and Announces DARPA Collaboration
Our portfolio company Medra has launched the AI Experimentalist, the scientific reasoning layer of its Physical AI Scientist Platform, and announced a collaboration funded by the Defense Advanced Research Projects Agency (DARPA).
The AI Experimentalist lets scientists set a high-level research goal and turns it into an executable experimental plan, running the full loop from literature review and experimental design through lab orchestration, data interpretation, and protocol refinement. It pairs with Medra's Physical AI Lab execution layer, using model-agnostic, multi-agent systems so the reasoning and the bench work operate as one platform with limited human oversight while researchers keep control over priorities.
Under the DARPA collaboration, Medra is developing the ability to translate natural-language objectives and human-written protocols into machine-executable experiments that can be measured, learned from, and improved over time. Partners access the system either through Physical AI Labs deployed on site or remotely through Medra Lab 001, the company's flagship autonomous facility that opened in April 2026 after just 77 days of construction. ML001 is now running closed-beta projects across academia, biopharma, and government, including DARPA, with assays in antibody discovery, protein engineering, gene editing, and cell biology.
Founder and CEO Michelle Lee has framed the bottleneck in modern science as validation: as AI models get better at proposing new molecules and hypotheses, the slow step becomes designing and running the experiments to test them. The AI Experimentalist is Medra's answer to that gap, building on its $52M Series A and public launch earlier this year.
