Remote Part-Time AWS Trainium NKI Kernel Engineer

Anyone AI Inc.; United States; Remote; CLP 81,919,000 - 163,837,000; Part time

Benefits offered by this job

Remote work

Part-time engagement

Focus on Trainium/NKI kernel tech

Job summary

Anyone AI Inc. is seeking experienced AWS Trainium/NKI kernel engineers to evaluate and improve kernel development for AI workloads. You will assess correctness, idiomatic usage, and performance on Trainium hardware.

Youll work remotely on a part-time, project-based engagement, reviewing CUDA→NKI migrations, memory hierarchies (SBUF, PSUM, HBM), tile-based computation, and DMA scheduling, providing clear written feedback.

Qualifications

• 2+ years hands-on experience developing or optimizing NKI kernels.
• Experience with AWS Trainium and/or Inferentia2.
• Strong understanding of tile-based computation, memory hierarchies (SBUF/PSUM/HBM), and DMA.
• Ability to evaluate CUDA→NKI migrations and optimize for Trainium.
• Excellent written feedback on complex kernel implementations.

Responsibilities

• Review and evaluate tasks involving NKI kernels and Trainium-specific patterns.
• Assess correctness and idiomatic optimization for Trainium hardware.
• Benchmark and profile workloads, identify bottlenecks and improvements.
• Provide clear written evaluations and recommendations.

Skills

NKI kernel development

Trainium optimization

CUDA migration evaluation

Profiling workloads

Technical feedback

Tools

AWS Neuron SDK

Neuron Compiler

CUDA

Triton

Job description

Anyone AI Inc. is seeking experienced AWS Trainium/NKI kernel engineers to evaluate and improve kernel development for AI workloads. You will assess correctness, idiomatic usage, and performance on Trainium hardware.

Youll work remotely on a part-time, project-based engagement, reviewing CUDA→NKI migrations, memory hierarchies (SBUF, PSUM, HBM), tile-based computation, and DMA scheduling, providing clear written feedback.