Applied Safety Research Engineer, Safeguards

This role sits at the intersection of applied ML research and engineering, focusing on how we measure and improve model safety.
Design experiments to generate representative test data, simulate realistic user behavior, and validate grader accuracy.
Investigate how factors like multi-turn conversations, tools, long context, and user diversity affect safety performance.
Productionize successful research into evaluation pipelines that run during training, launch, and beyond.
Collaborate with policy and enforcement to translate real-world harm patterns into measurable evaluations and user-facing tooling.
Ideal candidates have 4+ years of ML or software engineering experience, strong Python skills, data pipeline expertise, and a passion for AI safety.