Absci replaced spreadsheet sample handoffs with a Benchling registry and API integration to improve data for AI model training.

Absci's protein-therapeutics discovery relied on spreadsheet handoffs, with duplicates, incomplete datasets and copy/paste errors affecting AI model training. It implemented Benchling Bioresearch with a custom registry schema and API integration, including a formal request handoff data model that tracks samples in machine-readable form. The page reports better data capture and tracking, higher throughput across the screening funnel and better interoperability between teams, without numerical figures.

Benchling is a cloud R&D platform used by biotech and pharma for electronic notebooks, sample registry and structured scientific data.

This sits in the Build stage of the AI use-case lifecycle: Named examples from the foundation-model labs and platforms pharma teams use to actually construct, train and generate the model or application.