Member of Technical Staff, ML Systems

About the Role

Join an AI infrastructure team building and optimizing training and inference systems for image, video, diffusion, and world-model workloads. You will improve performance across GPU kernels, distributed engines, and multi-node systems, helping make demanding model workloads faster and more efficient.

What You'll Do

• Optimize training and inference performance across GPU hardware and distributed systems.

• Use Nsight and related tools to profile bottlenecks in kernels, memory, runtime systems, and clusters.

• Implement low-level optimizations using CUDA and Triton.

• Design and improve distributed training and inference engines for diffusion and multimodal models.

• Improve communication across GPUs and nodes using NCCL, RDMA, InfiniBand, or RoCE, including for disaggregated serving.

• Build benchmarks and regression tests to maintain performance improvements in production.

• Co-design kernels, runtimes, and models with hardware capabilities in mind.

What We're Looking For

• At least 2 years of experience in deep learning training or inference systems, distributed systems, or related ML systems work.

• Recent hands-on experience improving ML systems or GPU performance, including work on inference or training runtimes, kernels, or distributed execution.

• Experience authoring core features in an ML training or inference framework, such as vLLM, SGLang, TensorRT-LLM, or Megatron.

• Direct NVIDIA GPU kernel development experience with CUDA, CUTLASS, Triton, or PTX, and proficiency with PyTorch and Nsight.

• Strong computer science and machine learning fundamentals, with a degree in computer science or a related quantitative field.

• Experience optimizing diffusion, video, image, or other multimodal workloads is valuable.

Compensation & Benefits

Salary range: $180,000 to $230,000 annually. Visa sponsorship is available.

Location

On-site in Menlo Park, California, United States.