Senior AI Infrastructure Engineer, Physical Infrastructure

What you’d do

Own the stability of large-scale GPU training clusters, building self-healing systems to detect and isolate hardware faults and tuning NCCL, high-speed networking, Kubernetes, Run:AI, and Ray scheduling. Full-time role in Seattle, Washington. Found on the employer's careers page on Oct 8, 2026. Apply through the link.

You’ll need

• 10+ years in hands-on infrastructure, HPC, or datacenter engineering
• GPU compute at scale experience
• H200/B200/B300 GPU bring-up, cabling, firmware/driver management
• NVLink, InfiniBand, RoCE, or Spectrum-X interconnects
• Clusters of hundreds of GPUs