Periodic Labs · San Francisco
Data & AIEngineeringFullTime
About Periodic Labs
We're an AI and physical sciences company building state-of-the-art models to accelerate breakthroughs across materials, energy, and beyond. Backed by world-class investors and growing rapidly, we operate at the pace the frontier requires. Our team brings deep expertise, genuine ownership, and an insatiable drive to push the boundaries of what's scientifically possible.
About the Role
At Periodic Labs, we are automating scientific research in materials discovery; the data we collect, and how we represent it, is what makes this possible. We are hiring a materials scientist to work across the entire process: to find where our agents and data fail, and to fix those failures at the source. This is a hybrid research and infrastructure role. Roughly half your time will be spent working directly with our experimental and computational scientists, building agents and tools that solve active research problems. The other half will be spent improving the data those agents depend on.
This role requires domain depth, not just data engineering skill. You need to know, for example, what metadata actually matters for a given characterization technique. The research half of the role keeps you grounded in the problems we are actually using LLMs to solve, so the schemas you build serve real research rather than an abstraction of it. The work demands attention to detail and a willingness to get into the weeds: to meticulously read, understand, and improve our data. You'll be as much a materials scientist doing research as the person who makes our agentic harness actually work.
What You'll Do
Work directly with lab scientists and the computational team on active research problems, staying close to the actual bottlenecks LLM agents are meant to solve.
Investigate agent traces to identify data errors and agentic failure modes, and eliminate them at the source.
Restructure and re-architect our materials databases (lab experiments, characterization data, computations) around how LLM agents actually reason and fail, not just around human readability. Draw on domain expertise about how to represent the data and what metadata matters.
Work with the hardware and automation teams to make data collection more robust.
Build research software for lab environments, translating scientific requirements into working tools.
Distill what you learn into evaluations that measure agent performance.
Mechanics
Minimum experience: 4+ years of research or industry experience in experimental or computational materials science and working with data at scale.
Minimum education: PhD in Materials Science, Chemistry, or a related field, or equivalent industry experience.
Location: Menlo Park, CA (Soon: San Francisco, too)
Compensation: $250,000-350,000 + equity
Visa sponsorship: Yes, we sponsor visas.