Exscientia

Exscientia, now unified with Recursion, is an AI-driven drug discovery platform built on the Recursion Operating System (Recursion OS). It pairs one of the world's largest proprietary biological and chemical datasets—over 50 petabytes spanning phenomics, transcriptomics, proteomics, ADME, and de-identified patient data—with automated wet labs that use robotics and computer vision to run millions of cell experiments per week. Every result feeds back into the platform, training machine learning models to identify novel targets and design optimized molecules. This closed-loop system aims to cut the 90% failure rate of traditional drug discovery by grounding AI predictions in real experimental data. To process the immense data load, Recursion built BioHive-2 with NVIDIA, described as biopharma's most powerful supercomputer. The platform is used by pharma partners and internal programs to accelerate early-stage development, from hit identification through IND-enabling studies. The pipeline focuses on oncology (solid tumors, lymphomas, B-cell malignancies) and rare diseases (familial adenomatous polyposis, hypophosphatasia), with recent progress including the first patient dosed in the Phase 1 trial of REC-3565, a selective MALT1 inhibitor for B-cell lymphomas. Partnerships with Genentech and Roche support discovery and clinical validation—the Genentech collaboration recently unlocked its first neuroscience target, expanding beyond oncology into CNS disease. Compared to other AI drug discovery platforms, Exscientia's edge is its proprietary data scale and end-to-end wet lab integration. The closed-loop feedback between experiments and models is a genuine differentiator, and the Genentech and Roche partnerships add credibility. However, clinical validation is still emerging—the pipeline is early-stage, with no approved drugs yet. This platform suits organizations willing to commit to a data-rich, platform-driven approach rather than a lightweight, off-the-shelf tool.