jobs
Staff Software Engineer - GenAI Performance and Kernel
P-1285 About This Role As a staff software engineer for GenAI Performance and Kernel, you will own the design, implementation, optimization, and correctness of the high-performance GPU kernels powering our GenAI inference stack. You will le
- Company
- Databricks
- Location
- San Francisco, California
- Status
- Open
- Posted
- 2025-10-08T04:06:51+00:00
Databricks is hiring a staff software engineer to own the design, optimization and correctness of GPU kernels behind its GenAI inference stack. The work covers attention, MLP, softmax and layernorm kernels, plus vectorization, tiling, fusion, mixed precision, quantization and memory management, alongside profiling and verification tooling. The posting asks for hands-on kernel tuning in CUDA, Triton, OpenCL or assembly, GPU architecture knowledge, and experience with libraries such as CUTLASS or cuDNN. The role is on-site in San Francisco and includes mentoring and production rollout.
Original job posting