Staff Software Engineer, AI Safety
Core
Build shared software, tools, and systems to help researchers understand and reduce societal risks from scientific AI, including safety evaluations, red-teaming, and mitigation experiments.
Role type
Staff Software Engineer (AI Safety)
Builds
Reusable services, libraries, model/tool integrations, data pipelines, and applications for organizing evidence and inspecting results.
Domain
AI Safety / Scientific Computing
Deliverable
production ML models
Required skills
Python, backend engineering, API design, database management, data pipeline construction, system design, testing and debugging, translating researcher needs into tools, engineering judgment.
Preferred skills
AI research, AI safety, scientific computing, distributed experiment execution, model-serving integrations, isolated environments for agent evaluations.
Technologies
Python, APIs, databases
Responsibilities
Build reusable services and libraries for evaluations and red-teaming; develop applications for researchers to organize evidence and investigate failures; build data pipelines to preserve datasets and make findings traceable; own reliability and usability of shared safety tools; collaborate with core AI and software engineering teams to share tooling.
Seniority
Staff, hands-on IC
