GPU Infrastructure Engineer for ML Systems (Hybrid)

H Company; Greater London; Hybrid; GBP 85,000 - 120,000; Full time

Job summary

H Company is seeking an experienced infrastructure engineer to own the GPU compute powering our model training, evaluation and serving systems. You will operate the cluster stack from Kubernetes down to the GPU hardware, ensuring reliable production and optimizing performance for training and serving workloads.

You will run GPU-enabled Kubernetes clusters, improve scheduling and placement, and coordinate capacity with cloud providers.

Qualifications

• Have built and run GPU workloads on Kubernetes.
• Able to design GPU clusters for training and RL from scratch.
• Debug at OS level and below (Linux, drivers, networking, storage).
• Write strong Python.

Responsibilities

• Running Kubernetes clusters for GPU training and inference.
• Improving GPU scheduling and placement.
• Detecting, draining and repairing unhealthy GPU nodes.
• Managing capacity with cloud providers.
• Debugging GPU, Linux, storage and networking failures.

Skills

GPU workloads

Kubernetes

Python

System design

SRE experience

Tools

Kueue

Karpenter

AWS / EKS

Terraform

SkyPilot

Job description

H Company is seeking an experienced infrastructure engineer to own the GPU compute powering our model training, evaluation and serving systems. You will operate the cluster stack from Kubernetes down to the GPU hardware, ensuring reliable production and optimizing performance for training and serving workloads.

You will run GPU-enabled Kubernetes clusters, improve scheduling and placement, and coordinate capacity with cloud providers.