Research Scientist, Pre-training Research

We are hiring research scientists to work on foundation model pre-training. Our models learn the foundations of molecular structure and interaction, and their success depends on carefully designed architectural components, training objectives, and data mixes.

At Chai, pre-training research is largely domain-agnostic: understanding model architecture, reading training curves, and knowing when to move between phases of a recipe. We are biased toward simplification, generality, and scale. We adapt the best ideas from frontier language, image, and video models into our domain.

Our models have moved beyond protein structure prediction into real-world therapeutic engineering, and you will have the compute and the team to push that further.

We look for researchers with a strong track record, which can take many forms: getting a model to work, developing impactful theory, or making a substantial contribution to the field. In particular we look for: