At its core, Recursion operates the Recursion Operating System (Recursion OS) — an integrated platform combining proprietary multimodal biological datasets, machine learning models, and high-performance compute. The dataset exceeds 50 petabytes and spans phenomics, transcriptomics, proteomics, ADME profiling, and de-identified patient data, all generated through a highly automated wet lab that captures millions of cell experiments per week using robotics and computer vision.

Machine learning models trained on this data are used to rapidly identify novel disease targets and design optimized molecules, feeding back into a continuously improving discovery loop. To handle the computational demands at scale, Recursion partnered with NVIDIA to build BioHive-2, described as biopharma's most powerful supercomputer. The platform supports the full drug development lifecycle from hit identification through IND-enabling studies, demonstrating measurable improvements in speed, efficiency, and cost over traditional pharmaceutical averages.