Consultant – Operations Oversight & Performance, AI Infrastructure

About the role

The Consultant – Operations Oversight & Performance, AI Infrastructure will oversee the ongoing operational delivery of GPU‑based AI runtimes after go‑live. The role ensures Kubernetes clusters, GPU scheduling, container runtimes and autoscaling meet agreed performance standards while driving continuous improvement.

Key responsibilities

• Define, agree and monitor operational KPIs and SLAs with third‑party providers, covering availability, incident response, resolution times, change success rate and capacity utilization.
• Escort the hardening of Kubernetes clusters with NVIDIA GPU Operator, device plugins, DRA, MIG/time‑slicing and advanced schedulers such as Kueue.
• Review GPU‑aware scheduling, fair‑share quotas, priority classes, pre‑emption and multi‑tenancy isolation.
• Supervise container runtime, AI training/inference workloads and autoscaling mechanisms.
• Integrate observability (Prometheus/Grafana + DCGM), logging and GitOps for the runtime layer.
• Support multi‑cluster federation and lifecycle management in partnership with Rafay.
• Conduct regular operational reviews, drive service improvement plans and hold third‑party providers accountable.

Required profile

• 10+ years in GPU/HPC infrastructure operations, service management or managed‑service oversight.
• 7+ years of Kubernetes experience (CKA preferred) with large‑scale GPU workloads.
• Strong knowledge of the CNCF ecosystem, operators, CRDs and NVIDIA‑specific features.
• Excellent analytical, reporting and stakeholder management skills.

Required skills

• GPU infrastructure and high‑performance compute operations.
• Kubernetes, NVIDIA GPU Operator, DRA, MIG, Kueue (or equivalents).
• Observability tools: Prometheus, Grafana, DCGM.
• GitOps practices and multi‑cluster federation (Rafay).
• Container runtime management and AI workload orchestration.