Applied Research - Scientific AI

Role overview Build and post-train AI agents that can carry out scientific research tasks. The work combines autonomous research and computational science: agents may search literature, write and run code, use domain-specific tools, and select follow-up experiments. You will develop research environments and evaluations around partner tools, data, and experimental feedback, while contributing to a substantive research agenda.

Responsibilities - Develop and post-train agents for scientific research workflows. - Build or adapt environments and evaluations using research partners’ tools, data, and feedback. - Work directly with customers, users, or research partners while advancing research goals. - Create artifacts that other researchers and engineers can run, inspect, and extend. - Take ownership of ambiguous problems and keep collaborators informed as the work develops.

Requirements - Strong Python skills and the ability to produce usable, inspectable research artifacts. - Comfort working through open-ended problems and sharing progress with partners and teammates. - Evidence of strong work through publications, open-source systems, benchmarks, or shipped products; be prepared to explain what you built, what failed, and how evidence shaped your approach.

Benefits and work setup Full-time role with locations in San Francisco, New York, London, Berlin, or remote. Listed benefits include cash compensation of $150,000–$300,000 plus equity incentives, visa sponsorship and relocation support, a professional development budget, and team off-sites and conference attendance.