PhD in computer science, physics, mathematics, or a related field, or equivalent depth of experience gained through building AI systems at scale.
You have worked at a frontier AI research lab where research engineers are treated as first-class contributors. You know how to run large-scale experiments and you’ve been part of making that actually happen at scale.
You are comfortable across the full stack: distributed training frameworks, GPU/accelerator optimization, data pipelines, experiment tracking, and making research reproducible and reliable. You have experience down to the level of CUDA kernels or XLA.
You have contributed to open source projects like PyTorch, JAX, MLX or similar.
You care about research outcomes as much as system uptime. You form opinions about what experiments to run and how to design them. You can read papers and research reports and figure out what it would take to implement them as efficiently as possible.
A backround in mathematics or physics is strongly preferred. The best research engineers bring quantitative intuition to system design decisions.
You have strong software engineering practices: clean code, good testing habits, focus on performance, and an instinct for building systems that other people can actually use. In a small team everyone depends on what you build.
You are genuinely curious about biology. The data you’ll work with encodes billions of years of evolution, drawn from ecosystems most datasets never touch. You should find that interesting and not incidental.
Low ego, collaborative instincts, and a startup mentality. You’re comfortable with ambiguity and happy to wear multiple hats in a team where everyone contributes beyond their job description.