Senior DevOps Engineer (Full Remote from France)

We're hiring our first dedicated Devops / MLOps Engineer

Our ML team runs +100 active AI projects on GCP and needs the industrialized infrastructure to match. ML team is moving fast and they need a dedicated expert to connect the dots, align on the right practices, and help both sides deliver at their best. You'll be that person. You'll work at the heart of an AI-first environment, where shipping models fast and reliably is the core business, not a side project.

Your Mission

• Empower ML Engineers with the tools, infrastructure, and frameworks they need to iterate fast autonomously.
• Accelerate time-to-market for production-ready ML products: seamless integration, proper service connections, access to data and resources.
• Own ML CI/CD in close collaboration with the ML team, adapting existing frameworks to ML-specific needs, not just consuming them.
• Keep ML Engineers in control of their models in production: monitor, troubleshoot, iterate, refine directly in prod, no staging/prod mirror.
• Enable large-scale ML experimentation: robust, reproducible, scalable environments for both internal tests and A/B testing in production.
• Deliver concrete MLOps building blocks (MLflow, Kubeflow, KubeRay...) and manage GPU infrastructure dynamically, you've seen L40 shortages mid-training, you know how to handle them.
• Tackle technical debt on existing projects while laying the right foundations for what's next.
• Be the technical mediator between ML and Backbone teams understand both sides, propose solutions that stick.
• Handle run responsibilities: on-call, post-mortems, level-1 failure analysis.

What We're Looking For

• Solid MLOps or DevOps, background projects shipped in prod matter more than years on a resume.
• GCP expert: Vertex AI, GKE, GCS, BigQuery.
• Full GitOps: FluxCD first, ArgoCD accepted.
• Kubernetes in prod, not just in a lab.
• Hands-on with real MLOps tools: MLflow, Kubeflow, KubeRay.
• GPU-aware: you've managed GPU scarcity at scale during mass training runs.
• Python is a must, Bash expected, Go or Rust a plus.
• IaC (Terraform), containerization (Docker, Helm), observability (Prometheus, Datadog, Looker).
• AI Coding Assistants (Claude, Cursor, Dust)
• Data lifecycle management (cost, security, encryption)
• Fluent in French and English.

Nice to Have

• Jupyter Notebooks, broader ML/AI ecosystem
• Data pipelines (Airflow, Dataflow, Kestra)
• Redis clusters and infrastructure performance optimization

What Will Make the Difference

Beyond the tech stack, what will set you apart is your ability to act as a bridge rather than a silo. You thrive in ambiguity, stay calm under pressure, and turn complexity into clear solutions. You have a run culture, post-mortems and on-call sharpen you rather than drain you. And you lead by example: you advocate for best practices by winning people over, not by imposing them.