Founding Engineer, Environments + Data
Core
Design synthetic company environments, build graders and evaluation systems, and develop benchmarks to measure AI agent performance against real-world tasks.
Role type
Founding engineer specializing in AI agent evaluation environments and benchmarks
Builds
Synthetic evaluation environments, graders, and benchmarks for AI agents
Domain
Frontier AI / Agent Evaluation
Deliverable
production ML models
Required skills
Synthetic data pipelines, benchmark design, evaluation system implementation, reward hacking detection, software system simulation, model post-training
Preferred skills
High-level competition experience in mathematics, programming, debate, or games
Technologies
Synthetic data pipelines, evaluation systems, software simulation tools
Responsibilities
Design synthetic company environments with realistic data, build graders for advanced agents, develop benchmarks correlating with production performance, identify and fix evaluation failure modes, simulate software tools for agent interactions
Seniority
Founding, hands-on IC