Mirendil
Member of Technical Staff, Model Evaluation
$300K–$400K
San Francisco · On-site · Staff · Full-time · Posted Jun 24, 2026
About the job
Build evaluation infrastructure to measure frontier AI model behavior and help research teams determine whether models are improving.
What you'll do
- Design evaluation frameworks for model capabilities beyond standard benchmarks.
- Build automated evaluation pipelines, regression detection, and agent-assisted workflows for inspecting model behavior.
- Instrument training runs with observability tooling and work with post-training and RL teams to inform training decisions.
Pay
$300,000–$400,000 USD base salary
AI focus
Develop evaluation systems and workflows for frontier AI models, including capability measurement, regression detection, and analysis of training-related behavior changes.
Model-written summary · the original posting is authoritative