Built for the era of AI-driven science

Every published materials paper contains experimental knowledge—conditions, protocols, outcomes—buried in prose and files that no search engine can reach.

GenData extracts that knowledge automatically, then works with the researchers who ran the experiments to verify and complete it. The result is a structured, queryable corpus of experimental conditions across the literature, including the failure results that never made it to publication.

We're building the data infrastructure that makes decades of experimental research finally searchable: not by generating new experiments, but by unlocking the ones that already exist.