Chaidiscovery Research Scientist at Chai Discovery improving real-world performance of foundation models for drug discovery. Responsibilities include data generation planning, evaluation design, and post-training recipe formulation.
Responsibilities
We are hiring research scientists to improve the real-world performance of our foundation models. The work involves planning targeted data generation campaigns, designing evals that accurately measure model performance, and formulating post-training recipes aimed at making models useful for real drug discovery problems.
We prioritize simple, robust methods over intricate solutions built for narrow problems. We particularly value researchers who can formalize messy biological problems using methods from other fields.
Qualification
Strong quantitative background (CSDeep software engineering expertiseWe are happy to bet on rising talent
Required
We look for technical depth paired with interest in the underlying biology. Biological background is not required. All of the following are positive signals:
Published research in machine learning for protein, antibody, or small-molecule modelling.
Fluency in the technical details of protein language, structure prediction, and design models (AlphaFold, RFDiffusion, ProteinMPNN, ESM, etc).
Strong quantitative background (CS, math, or physics), particularly in probability, statistics, and machine learning.
Deep software engineering expertise, with experience working in large codebases.