Recursion

Recursion – Technical Documentation Notes

Recursion's platform, Recursion OS, is the primary technical artifact of interest for integration and partnership contexts. The following structural components are documented across Recursion's public technical disclosures:

Platform Components

• Biological Experimentation Layer: Automated wet-lab systems capable of executing up to 2 million experiments per week. Robotics handle assay preparation, imaging, and sample management at scale.
• Data Infrastructure: A proprietary dataset exceeding 50 petabytes, encompassing cellular imaging, genomic, and chemical data. This corpus underpins all downstream ML model training.
• AI/ML Models: Trained on the proprietary dataset to perform tasks including target identification, compound prioritization, and generative de novo chemistry (novel molecule design).
• Dry Lab Integration: Computational pipelines are tightly coupled with wet-lab outputs, forming a continuous feedback loop between experimentation and model refinement.

Key Technical Differentiators

Capability	Detail
Weekly experiment capacity	Up to 2,000,000
Proprietary data volume	50+ petabytes
Compound reduction vs. industry	~90% fewer required
Chemistry approach	Generative de novo (novel structure design)
Lab integration	Fully integrated wet + dry automation

Pipeline

Recursion's internal pipeline is concentrated in oncology and rare disease, with programs at clinical stages.