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

THE OPPORTUNITY

A role worth
a closer look.

Use ML infrastructure experience to develop and assess technical tasks for frontier-model training. The work centres on kernels, accelerator diagnostics and inference performance, including evaluation rubrics and written solutions. This is an expert training-and-evaluation engagement, rather than a promise to operate a lab's production serving stack.

Experience highlighted by the employer

• At least two years in ML infrastructure, accelerator performance or model serving; practical depth in at least one listed systems area.
• JAX or PyTorch experience, clear technical writing and availability for 40 weekday hours without conflicting engagements.

Location & eligibility

The description explicitly permits worldwide remote candidates, with engagement through Cincinnatus or a local employer-of-record partner. Country availability and authorization still need confirmation. The header says hourly contract; the body describes a full-time engagement and generic contractor terms. Confirm the actual agreement, exclusivity and hours. The source says it cannot support H1-B or STEM OPT candidates.

Compensation

$90–$120 / hour · Confirm currency. Published source range uses a dollar sign without naming the currency. This is not a guaranteed offer; confirm currency and country-specific engagement terms.

Before you apply

This is TaskBull’s concise summary, not the employer’s full job description. Review the complete responsibilities, terms, accessibility information and privacy notice on the original application page.