GPU Platform Reliability Engineer - 4-Day Week, Remote

Hamilton Barnes; United Kingdom; Remote; GBP 108,000 - 132,000; Full time

Benefits offered by this job

4-day work week

Remote-first with flexibility

Uncapped holiday

Competitive compensation

Job summary

Hamilton Barnes is seeking a GPU Platform Reliability Engineer to own reliability, performance, and automation for the AI/GPU platform. You will bridge hardware telemetry with cloud-native orchestration, acting as the escalation point for a live GPU fleet and building automated remediation tooling.

The role includes a 4-day week, remote flexibility, and a competitive package. Expertise in NVIDIA stacks, Kubernetes and IaC is essential for success.

Responsibilities

• Act as the top internal escalation point for hardware, PCIe, GPU memory (ECC) and XID driver faults across the fleet.
• Build and maintain automated health check frameworks using NVIDIA DCGM, NCCL test suites and telemetry agents.
• Own automated rollout pipelines for GPU BIOS, system BIOS, BMC firmware and CUDA drivers across live clusters with no unannounced customer impact.
• Diagnose bottlenecks, link degradation and optimize data paths across InfiniBand and RoCEv2 fabrics to prevent GPU starvation.
• Tune NVIDIA GPU Operator, OS and workload plugins; design node-cordon, eviction and repair workflows from telemetry.
• Drive full-stack IaC (Terraform, Ansible, Python) for fleet expansion and incident war-room resolution with blameless post-mortems.

Job description

Hamilton Barnes is seeking a GPU Platform Reliability Engineer to own reliability, performance, and automation for the AI/GPU platform. You will bridge hardware telemetry with cloud-native orchestration, acting as the escalation point for a live GPU fleet and building automated remediation tooling.

The role includes a 4-day week, remote flexibility, and a competitive package. Expertise in NVIDIA stacks, Kubernetes and IaC is essential for success.