Home
Mercor

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

  • Pay: $90 to $120 / hour
  • Contract: Contract
  • Who can apply: Open worldwide
  • Posted Oct 22, 2026

In short

Earn 90 to 120 USD per hour from anywhere in the world. You'll work on the engines behind large language models with a leading AI lab's GenAI team. If you already build ML systems and love fast, efficient code, this remote freelance gig was made for you.

What you will do

  • Design tough challenges on kernels, profiling, debugging and inference serving, including problems that trip up top models, like Pallas kernels on TPU.
  • Write accurate, well-structured solutions to those challenges.
  • Review ML systems tasks and solutions, and give written technical feedback that experts can trust.
  • Build guidelines and detailed rubrics on kernel optimization, profiler output and serving speed trade-offs.

What you need to apply

  • 2+ years of hands-on experience in ML systems, ML infrastructure, model serving or accelerator performance.
  • Practical skills in at least one area: custom kernels, profiling and trace analysis, debugging distributed workloads, or serving LLMs at scale.
  • Working production experience with JAX and/or PyTorch.
  • Strong written communication, and the ability to work at least 40 hours per week on weekdays.

Who can apply

Open to candidates worldwide, as this role is fully remote. Your contract will be arranged through Cincinnatus LLC or one of its employer-of-record partners in your country.

A tip from Romowork

Lead your application with a concrete kernel, profiling or serving project you have shipped, and explain your choices clearly. We know clear writing matters a lot here, so show it in your application too.

Originally published on Mercor

Related jobs