Staff+ Fullstack Software Engineer, Safeguards Engineering
Core
Build safety and oversight mechanisms for AI systems, including internal review tools for analysts, interfaces for human supervision of agents, and visualization tools to surface abuse patterns.
Role type
Staff+ Fullstack Software Engineer (Safeguards)
Builds
Internal review tools, dashboards, case queues, evidence viewers, and human-agent supervision interfaces.
Domain
AI Safety / Trust & Safety / Software Engineering
Deliverable
production ML models
Required skills
TypeScript, React, Python, fullstack development, data-dense internal tooling, abuse detection, adversarial input mitigation, prompt engineering
Preferred skills
Experience with integrity/spam/fraud detection, AI/ML trust and safety mechanisms, working with operational teams
Technologies
TypeScript, React, Python
Responsibilities
Design and build internal review tools for analysts; build interfaces for human supervision of agents; surface abuse patterns via visualization tools; own frontend architecture for tools handling sensitive data at scale.
Seniority
Staff+, hands-on IC
