Remote GPU Kernel Evaluator & Optimizer

Appsierra Group; United States; On-site; USD 146,000 - 187,000; Full time

Benefits offered by this job

Fully remote

Flexible schedule

Weekly payments via Stripe or Wise

Job summary

Appsierra Group is seeking a GPU Kernel Expert for a fully remote, independent contractor position in the United States. You'll apply GPU kernel development and optimization expertise to evaluate kernel tasks used to train and assess frontier AI models, focusing on correctness, performance, and technical rigor.

You will review numerical tolerances, reference implementations, and run-time behavior across hardware and software environments.

Qualifications

• 3+ years hands-on GPU/kernel development, optimization, or verification.
• Experience with at least two of: CUDA, Triton, NKI, Pallas (JAX).
• Strong understanding of numerical correctness for GPU kernels: tolerances, ULP, reference implementations.

Responsibilities

• Evaluate GPU/accelerator kernels for quality, correctness, completeness, and rigor.
• Assess numerical correctness using absolute, relative tolerances and ULP criteria.
• Evaluate whether reference implementations are appropriate for validating kernel outputs.
• Review performance benchmarks for fairness, consistency, and methodological soundness.
• Assess task scope and determine whether requirements are clear, realistic, and well-defined.
• Evaluate compilation and runtime behavior across different hardware/software environments.
• Identify common failure modes (e.g., driver mismatches, OOM, shape/stride issues, autotuning failures).
• Provide structured, rubric-based written feedback.
• Evaluate kernel tasks involving generation, translation/lowering, migration, debugging, optimization, and operator fusion.

Skills

GPU kernel development

Performance profiling

Analytical thinking

Technical writing

Tools

CUDA

Triton

NKI

Pallas (JAX)

Job description

Appsierra Group is seeking a GPU Kernel Expert for a fully remote, independent contractor position in the United States. You'll apply GPU kernel development and optimization expertise to evaluate kernel tasks used to train and assess frontier AI models, focusing on correctness, performance, and technical rigor.

You will review numerical tolerances, reference implementations, and run-time behavior across hardware and software environments.