Hybrid GPU Infra Engineer — ML Training & Scheduling Expert

H Company; Paris; Hybrid; EUR 90,000 - 130,000; Full time

Benefits offered by this job

Hybrid work model

Competitive compensation package

Job summary

H Company in Paris is seeking an experienced Infrastructure Engineer to own GPU compute for model training, evaluation and serving systems. You will run Kubernetes clusters for GPU workloads, design scheduling strategies, and manage capacity across cloud providers.

You’ll work with the Platform Infrastructure team and research/product teams, collaborating with cloud providers. Strong Python and Linux debugging skills are essential.

Qualifications

• Experience running GPU workloads on Kubernetes in production environments.
• Ability to design GPU clusters for training and onlineRL from scratch.
• Strong debugging skills across Linux, GPU drivers, networking and storage.
• Proficiency in Python for tooling and automation.
• Experience in capacity planning and cloud-provider integration.

Responsibilities

• Run Kubernetes clusters for GPU training and inference.
• Improve GPU work scheduling and placement.
• Detect, drain and repair unhealthy GPU nodes.
• Manage capacity with cloud providers.
• Debug failures in GPUs, Linux, storage and networking.

Skills

GPU workloads

Scheduling & placement

Linux debugging

Python

Cloud capacity management

Tools

Slurm

Kueue

Karpenter

SkyPilot

AWS SageMaker HyperPod

NCCL

Terraform

Open MPI

Job description

H Company in Paris is seeking an experienced Infrastructure Engineer to own GPU compute for model training, evaluation and serving systems. You will run Kubernetes clusters for GPU workloads, design scheduling strategies, and manage capacity across cloud providers.

You’ll work with the Platform Infrastructure team and research/product teams, collaborating with cloud providers. Strong Python and Linux debugging skills are essential.