• Own the client's recommendation system end-to-end: design, training, deployment, and monitoring • Build and mature ranking and retrieval models that process behavioral signals — clicks, opens, purchases — at consumer scale • Operate ML systems at hundreds of millions of events per day, with direct impact on client revenue • Make independent technical decisions and engage directly with client engineering leadership — no intermediaries

• Design and implement multi-agent AI architectures using AWS Bedrock • Develop agent orchestration logic and collaborative agent workflows • Configure and manage AWS Bedrock Agents, Knowledge Bases, and Guardrails • Build Retrieval-Augmented Generation (RAG) pipelines using vector databases and embeddings • Implement tool integrations using Model Context Protocol (MCP) and API-based services • Optimize LLM behavior through prompt engineering, tuning, and context management • Develop observability and monitoring strategies for AI workflows using CloudWatch and X-Ray • Build scalable event-driven architectures and resilient integration patterns • Design error handling, retry strategies, and graceful degradation mechanisms • Collaborate with engineering, product, and architecture teams to deliver production-grade AI solutions • Support infrastructure automation and deployment pipelines using IaC and CI/CD practices • Ensure governance, security, auditability, and compliance standards across AI systems

• Design, build, and ship LLM-powered features and agentic workflows that serve real Gametime users in production. • Build and maintain evaluation frameworks, prompt testing pipelines, and regression suites that ensure quality and reliability of AI-powered experiences. • Contribute to the orchestration layer, including agent routing, tool use, state management, and multi-step workflow coordination. • Develop and optimize prompt optimization strategies, structured outputs, and LLM integration patterns across the platform. • Propose architecture decisions and technical designs for review by the team's tech lead, balancing speed with long-term maintainability. • Collaborate cross-functionally with product, engineering, and data teams to translate customer needs into AI system design. • Stay current with the rapidly evolving LLM and agentic AI landscape, bringing practical new techniques into the team's toolkit.

Role Description We're hiring our first dedicated MLOps Engineer. Our ML team runs +100 active AI projects on GCP and needs the industrialized infrastructure to match. ML and DevExp teams are moving fast on their respective fronts 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 DevExp 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 DevExp: understand both sides, propose solutions that stick. - Handle run responsibilities: on-call, post-mortems, level-1 failure analysis. Qualifications - 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). - Fluent in French and English. Requirements - Jupyter Notebooks, broader ML/AI ecosystem. - AI Coding Assistants (Claude, Cursor, Dust). - Data pipelines (Airflow, Dataflow, Kestra). - Data lifecycle management (cost, security, encryption). - Redis clusters and infrastructure performance optimization. Benefits - Don't check every box? Apply anyway. - We're looking for the right person, not the perfect resume. If this role excites you, let's talk. - Dailymotion is an equal opportunity employer. All positions are open to people with disabilities. Need accommodations? Just let us know. Interview Process - HR Interview with Marvin (30mn) - Manager interview with Cédric (1h) - Technical Interview with 2 Architects (1h): no coding test, just a real technical conversation. - ML Interview with Brice & Samuel (1h) - Final Interview with Alan, VP Platform (1h) Additional Information - Type of Contract: Permanent - Department: Operations - Compensation: EUR 75000 - EUR 85000 yearly