ML Inference Engineer â Terminal-Bench Task Designer (vLLM / SGLang / TensorRT-LLM)
Core
Design expert-level Terminal-Bench tasks by converting real ML inference work into reproducible, command-line solvable challenges for AI agents.
Role type
Senior IC ML Inference Engineer (Task Designer)
Builds
Expert-level benchmark tasks with automated checks, Docker packaging, and reference solutions
Domain
AI Agents / ML Inference / Benchmarking
Deliverable
production ML models
Required skills
ML inference and serving, vLLM, SGLang, TensorRT-LLM, PyTorch internals, JAX, Python, command line, Docker, Git, automated testing, numerical drift detection
Preferred skills
AI lab or LLM infrastructure, PyTorch/vLLM/SGLang contributions, GPU profiling
Responsibilities
Design expert-level Terminal-Bench tasks from real ML inference experience, define files/systems/environments for realistic challenges, write working reference solutions, build automated checks for correctness, package tasks with Docker and Git, iterate on tasks based on feedback
Seniority
Senior, hands-on IC
