Founding Engineer, Environments + Data
Core
Design synthetic company environments, build robust graders and evaluation systems, and develop benchmarks to measure AI agent performance against realistic work tasks.
Role type
Founding Engineer (AI Agent Evaluation & Environments)
Builds
Synthetic evaluation environments, benchmarks, and grading systems for AI agents
Domain
Artificial Intelligence / Agent Evaluation / Software Simulation
Deliverable
production ML models
Required skills
AI agent evaluation, synthetic environment design, grader development, benchmark creation, reward hacking detection, software system simulation
Preferred skills
Frontier AI lab experience, early-stage startup experience, competitive problem-solving in math/programming/debate/games
Technologies
AI agent frameworks, simulation tools, evaluation pipelines
Responsibilities
Design synthetic company environments with realistic data, build robust graders and evaluation systems, develop production-ready benchmarks, identify and fix evaluation failures like reward hacking, simulate software tools for agent interactions