Quick overview
Job description
About the Role
Join a team building the training and inference stack for image, video, and world-model workloads. You will improve performance across GPU kernels, distributed execution, and serving systems.
What You'll Do
Optimize training and inference performance for image, video, and world-model workloads.
Profile kernel, memory, system, and cluster bottlenecks using Nsight and related tools.
Implement low-level GPU optimizations with CUDA and Triton.
Design and improve distributed training and inference engines for diffusion models.
Improve communication performance across GPUs and nodes, including NCCL, RDMA, InfiniBand, RoCE, and disaggregated serving.
Build benchmarks and regression tests to sustain performance improvements in production.
Apply hardware-aware co-design across kernels, runtimes, and models.
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.
At least 1 year of recent, hands-on experience with ML systems or GPU performance.
Experience authoring core features in an inference or training framework, rather than focusing only on deployment or integration.
Hands-on experience writing GPU kernels for NVIDIA hardware using CUDA, CUTLASS, Triton, or PTX.
Strong machine learning and computer science fundamentals, with a degree in computer science or a related quantitative field.
Experience optimizing inference or training performance, including GPU kernels, runtimes, or distributed execution.
Experience with diffusion, video, image, or other multimodal workloads is valuable.
Compensation & Benefits
Base salary: $180,000 to $260,000 annually. Visa sponsorship is available.
Location
On-site in Menlo Park, California, United States.
Haystack's AI agent can fill in and submit this application for you. How AI auto apply works