Medra launched AI Experimentalist, a reasoning layer for its physical AI drug discovery robotics platform, in a collaboration with DARPA. The system translates natural-language research goals into executable wet-lab and analysis workflows spanning literature review, experimental execution, and protocol refinement. Medra said it can run optimized assay cascades with iterative feedback, potentially compressing time-to-data from days to hours by continuously feeding results into subsequent runs. The company also highlighted that its approach is meant to address a core bottleneck—automating experimental validation at scale rather than only generating hypotheses. The announcement signals continued defense-linked investment in agentic lab automation, with an emphasis on physically executing and learning from experiments end-to-end.