Technical Product Manager - AI Infra Resilience
Core
Own the roadmap for workload failure attribution in GPU clusters, determining if failures stem from hardware, system software, or application code and surfacing verdicts to operators and developers.
Role type
Senior Product Manager (AI Infrastructure)
Builds
NVIDIA DSX Platform (open, modular infrastructure software for AI factories)
Domain
AI Infrastructure / GPU Clusters / Distributed Systems
Deliverable
production ML models
Required skills
GPU failure triage, XID codes, DCGM metrics, NCCL traces, distributed training failure patterns, open-source community engagement, cloud partner collaboration
Preferred skills
SRE or infra engineer background on GPU clusters, GPU scheduling (Slurm, Kueue, Kubernetes), observability APIs at data center scale
Technologies
DCGM, XID, NCCL, NVLink, PCIe, Slurm, Kueue, Kubernetes
Responsibilities
Drive product vision and roadmap for failure attribution; Identify gaps in customer triage processes; Work with cloud partners to translate operational requirements; Drive community engagement for open-source attribution projects
Seniority
Senior, hands-on IC
