Google Technology/Software

Research Software Engineer, Protein Design, DeepMind

London, United Kingdom In person Full-time

Identify and solve bottlenecks. Figure out what's slowing the team down and fix it, this means understanding how the research loop works end-to-end from model training through design generation, in silico evaluation, and wet-lab validation and deciding where engineering effort will have the most leverage. Own the inference and evaluation stack. Maintain and improve our design pipeline, evaluation frameworks, and the data layer connecting them.

Make it easy to move fast, the team iterates quickly on research code. Make that sustainable — improve test coverage where it matters, simplify dependency management, reduce build times, and make the codebase easy to work in without creating process overhead. Collaborate directly with ML, biology, and chemistry researchers to identify and independently solve technical hurdles, contributing your own research as opportunities arise.

Job description provided by Google.
View the original posting