GPU Kernel Quality Engineer

Obsidian; New York (NY); Remote; USD 140,000 - 200,000; Full time

Job summary

Obsidian seeks a detail-focused evaluator to assess GPU/accelerator kernel development tasks used to train and evaluate frontier AI models. You will review numerical correctness, benchmark fairness, task scoping, and ensure compilation and runtime validity across diverse kernel types.

The role emphasizes rubric-based written feedback, reproducible methodology, and collaboration with ML scientists to tighten kernel pipelines.

Qualifications

• 3+ years of hands-on experience developing, optimizing, or verifying GPU/accelerator kernels in at least two of: CUDA, Triton, NKI, or Pallas (JAX)

Skills

GPU kernels

CUDA

Triton

Performance profiling

Numerical correctness

Tools

Nsight

NVIDIA NCU

Roofline analysis

Job description

Obsidian seeks a detail-focused evaluator to assess GPU/accelerator kernel development tasks used to train and evaluate frontier AI models. You will review numerical correctness, benchmark fairness, task scoping, and ensure compilation and runtime validity across diverse kernel types.

The role emphasizes rubric-based written feedback, reproducible methodology, and collaboration with ML scientists to tighten kernel pipelines.