Enterprise Data Science in Drug Development

In 2015, Bristol Myers Squibb had fragmented analytics -- siloed compute access, locally stored data, minimal version control, and roughly eight months to deploy an application, with no shared path from exploration to production. BMS standardized on Domino, scaling to 5,000 users across nine business units running 30,000 projects, via three phases: unrestricted sandboxing, scientifically rigorous output validation, and converting local wins into reusable templates. The platform helped BMS grow to support thousands of researchers and over 80 peer-reviewed publications; robotic screening combined with ML compressed months of manual chemistry experiments into a single week, and an internal optical chemical-structure-recognition dataset became an external benchmark OpenAI used to test next-generation models. Jacob Albrecht of BMS said 'true scale is about multiplying decisions rather than just adding users or compute.'

Domino Data Lab is an enterprise MLOps platform built for regulated industries — the sanctioned sandbox where biopharma teams run reproducible experiments under GxP/21 CFR Part 11 guardrails.

This sits in the Design & Stress-Test stage of the AI use-case lifecycle: Named examples from the tools pharma teams use to scope, prototype and benchmark an AI use case before committing to build it — data-catalog, sandbox and evaluation platforms.