ML Ops Engineer

Responsibilities

• Design, maintain, and evolve CI/CD pipelines using Azure DevOps for Brain APIs
• Package applications via Helm and manage Kubernetes configurations (requests/limits, autoscaling, secrets)
• Deploy and operate LLMs, including versioning, lifecycle management, and A/B testing
• Optimize model inference using runtimes like vLLM, TGI, or Triton
• Integrate LLM ecosystem components such as vector stores and orchestrators like LangChain
• Implement observability stacks (logs, metrics, traces) and SRE indicators (SLO, SLI)
• Act as the technical interface between the Brain team and the Group IT department regarding infrastructure and security
• Optimize infrastructure and inference costs through FinOps practices

Requirements

• Proficiency with ML/DL frameworks such as PyTorch or TensorFlow
• Experience with model serving tools like Triton, TorchServe, or KServe
• Strong knowledge of CI/CD (Azure DevOps) and GitOps (ArgoCD or Flux)
• Expertise in containerization with Docker and orchestration with Kubernetes
• Experience with Infrastructure as Code (Terraform, Ansible)
• Familiarity with observability tools like Prometheus, Grafana, or OpenTelemetry
• Ability to communicate technical AI constraints to IT teams
• High degree of autonomy and technical curiosity regarding LLM ecosystems

About the Company

Septeo is a leading software player in Europe, recognized among the top 5 sectoral software publishers in France. The group places artificial intelligence at the heart of its evolution, aiming to "make intelligent software for Europe."