Helix AI Engineer, Training Performance at Figure
San Jose, CA
$200,000–$400,000from the description
Aug 13, 2026
San Jose, CA
Nov 9, 2026
What this job asks for AI summary
An AI Training Performance Engineer role at a humanoid robotics company focused on pushing the limits of large-scale distributed model training. The work centers on optimizing GPU kernels, profiling and eliminating performance bottlenecks across 100k+ GPU clusters, evaluating novel accelerator hardware, and co-designing model architectures with researchers to maximize hardware utilization. Suited to engineers with deep GPU systems expertise and a track record of measurable performance gains on large models.
Senior level · 3+ years · San Jose-Sunnyvale-Santa Clara, CA · Bachelor's required · Full-time
“or” means any one of them counts — you don't need all of them.
We read this from the posting text with AI. Skim the description below before ruling yourself out.
How this req sits in the market our data
Roughly 80 people in the San Jose-Sunnyvale-Santa Clara, CA area plausibly meet what this posting asks for (software developers). range 25–120
Applicant volume Moderate — A normal amount of company. The rare requirements below are what will separate a shortlisted application from the rest.
Rare in this occupation — lead with these, and say what you built with them.
Most people in this occupation already list these. Still required — just not what gets you shortlisted.
What the occupation pays Median $217,796 (middle half $177,469–$231,052). This posting is about at that midpoint.
Estimated from BLS employment for this occupation and area, per-skill prevalence across our listing corpus, and published wage benchmarks — as of Aug 15, 2026. It is a model, not a headcount.
Why we read it this way (5)
The role sits at the intersection of systems software engineering and ML infrastructure; 15-1299 (Computer Occupations, All Other) is a plausible alternative given the highly specialized AI performance engineering focus, but 15-1252 (Software Developers) was chosen because the primary day-to-day work is writing and shipping code — custom kernels, tooling, and framework extensions.
Triton is listed in both the responsibilities (kernel writing) and requirements sections; it is captured as required. The Gluon compiler reference in responsibilities is treated as context rather than a standalone required skill, as it is not named in the Requirements block.
FSDP, context parallel, and expert parallel are listed together in the Requirements block as parallelism strategies to explore; FSDP is captured as the canonical representative with the others noted as context.
The 3+ years requirement is stated at the role level ('3+ years in AI performance engineering'), so it is reflected in the overall years minimum rather than attached to any individual skill.
The wide compensation range ($200K–$400K) is stated explicitly in the posting and reproduced exactly; the posting notes it may vary based on individual factors.
Read the full posting
The employer publishes the full description on their own site — read it there ↗. Or sign in to read it here — it's free, and it also lets you track this application.