The role involves developing foundational infrastructure for safety, oversight, and intervention mechanisms within AI systems. Engineers will build systems designed to detect unwanted model behaviors and prevent disallowed use. The work includes creating data storage, management, metric, and evaluation systems, as well as tooling for human and agentic review. Responsibilities also include maintaining operational systems to ensure safety and user well-being while reducing the need for manual intervention. The position requires building multi-layered defenses that function effectively at scale.

Candidates are required to hold a Bachelor’s degree in Computer Science, Software Engineering, or possess comparable experience. The role necessitates four to ten years of professional software engineering experience. Proficiency in Python and the ability to work across the stack are essential requirements. Additionally, candidates must demonstrate strong communication skills to effectively explain complex technical concepts to non-technical stakeholders.