Software Engineer, GPU Infrastructure
Core
Build and operate multi-cloud GPU/TPU Kubernetes clusters to train, evaluate, and serve foundation AI models.
Role type
Senior IC GPU infrastructure engineer
Builds
Shared Kubernetes fleets for training and serving frontier AI models
Domain
Cloud infrastructure + AI/ML hardware
Deliverable
infrastructure
Required skills
GPU Kubernetes cluster management, Linux debugging, networking (RDMA, NCCL), Python, Go, automation, fault detection/remediation
Preferred skills
Hardware vendor partnership, high-speed interconnects, fleet utilization optimization
Technologies
Kubernetes, Python, Go, RDMA, NCCL, Linux, multi-cloud environments
Responsibilities
Deploy and operate GPU Kubernetes clusters across clouds, automate node/GPU/fabric fault detection and remediation, troubleshoot production failures and performance problems, improve fleet utilization and cost tracking, partner with cloud providers and hardware vendors on reliability and performance
Seniority
Senior, hands-on IC
