Senior AI Training Infrastructure Engineer

Genesis; San Carlos (CA); On-site; USD 180,000 - 240,000; Full time

Job summary

Genesis is seeking an experienced engineer to optimize large-scale ML training systems. You will profile bottlenecks from data pipelines to GPU kernels, design scalable distributed training with PyTorch, and implement low-level CUDA/CuDNN/Triton optimizations that integrate into high-level training workflows.

Your work will ensure robust, high-utilization performance across multi-node GPU clusters, with emphasis on hardware-aware software optimization and tooling for rapid diagnosis of

Qualifications

• 8+ years in distributed ML infrastructure, HPC, or similar

Responsibilities

• Profile and reduce wall-clock time to convergence across the training stack.
• Design and optimize distributed PyTorch training on multi-node GPU clusters.
• Implement low-level CUDA/cuDNN/Triton kernels and integrate with high-level frameworks.
• Optimize CPU/GPU balance, memory, throughput, and networking for hardware efficiency.
• Develop monitoring and debugging tools for large-scale runs to diagnose regressions.

Skills

Python programming

Distributed systems

Performance optimization

System-level tuning

Tools

PyTorch

CUDA

cuDNN

Triton

Job description

Genesis is seeking an experienced engineer to optimize large-scale ML training systems. You will profile bottlenecks from data pipelines to GPU kernels, design scalable distributed training with PyTorch, and implement low-level CUDA/CuDNN/Triton optimizations that integrate into high-level training workflows.

Your work will ensure robust, high-utilization performance across multi-node GPU clusters, with emphasis on hardware-aware software optimization and tooling for rapid diagnosis of