MLOps Engineer, LLM Performance (Kernels, Profiling, Serving)

ML systems engineers write adversarial tasks and solutions on accelerator kernels, profiling, debugging and inference serving, and review them with written feedback and rubrics. The output becomes training data for frontier models. Full-time and exclusive, open to candidates worldwide.

Pay

$90–$120/hr

Level

Professional (working in the field)

Requirements

2+ years in ML systems or infrastructure, with hands-on kernels, profiling, distributed debugging or LLM serving; JAX or PyTorch in production

Hours

40 h/week

Field

Coding

Platform

Mercor