Postdoctoral Research Associate (Artificial Intelligence for Programmable Cloud Labs)

We aim to develop AI systems that can interact safely and reliably with the physical world—particularly scientific instruments—and help automate experimental research.

The successful candidate will work on topics including:
1) Harness engineering for scientific AI agents
2) Curating and evaluating reusable agent skills
3) Connecting coding agents with instruments and laboratory systems
4) Designing safe, auditable systems for autonomous experimentation
4) Developing domain-specific languages for scientific workflows

We welcome candidates with a PhD in engineering, computer science, applied mathematics, mathematics, physics, materials science, or a related scientific field. Candidates should have experience with agentic AI, particularly coding agents. Experience with harness engineering, tool use, agent evaluation, scientific automation, or laboratory instrumentation would be especially valuable. An exceptional qualification would be knowledge of category theory and an interest in using it to design principled domain-specific languages for scientific and engineering applications.

This position offers the opportunity to work in a highly collaborative environment spanning AI, experimentation, instrumentation, data, and physical modeling—and to help define how trustworthy AI agents can participate in real scientific laboratories.