Senior GPU Platform Engineer (Kubernetes + MLOps)

Ubisoft Paris Studio; Paris; Sur place; EUR 90 000 - 140 000; Plein temps

Avantages offerts par ce poste

Internal learning platform

Game library access

CSE discounts

Résumé du poste

Ubisoft Paris Studio is seeking a Platform/DevOps Engineer to lead GPU workloads and high-performance model serving. You will design, build and operate a GPU platform on GCP, implementing infrastructure-as-code and self-service compute for data science teams.

You will guide Kubernetes deployments, establish best practices, manage costs, and mentor future hires while collaborating with data scientists to move from PoC to production.

Qualifications

• Extensive hands-on experience with Kubernetes, Terraform and GitOps practices.
• Proficiency with GCP; AWS exposure is a strong plus.
• Platform-as-a-product mindset and ability to document solutions and set precedents.
• Fluent in English; able to communicate across diverse teams.

Responsabilités

• Design, build and operate a GPU platform on GCP for AI workloads as infrastructure-as-code.
• Deliver self-service compute for data science, including Ray on Kubernetes with quotas and cost visibility.
• Establish golden paths for deploying services to streamline production handoffs.
• Oversee deployment, hosting and operation of models in production environments.
• Control infrastructure costs including GPUs and related cloud resources.
• Collaborate with data science teams from PoC to production deployments.
• Continuously improve performance, reliability and efficiency; automate workflows.

Connaissances

Kubernetes

Terraform

GitOps

CRDs & operators

ArgoCD

Helm

GCP

AWS

Distributed Systems

DevOps

MLOps

Outils

KubeRay

Slurm

Kubeflow

Ray on Kubernetes

MLflow

Weights & Biases

Model registries

Description du poste

Ubisoft Paris Studio is seeking a Platform/DevOps Engineer to lead GPU workloads and high-performance model serving. You will design, build and operate a GPU platform on GCP, implementing infrastructure-as-code and self-service compute for data science teams.

You will guide Kubernetes deployments, establish best practices, manage costs, and mentor future hires while collaborating with data scientists to move from PoC to production.