MLOps

About

The SI Data Platform & BI domain wishes to strengthen its capabilities for industrializing Data Science and AI projects. The Dataiku DSS platform is currently used for developing Machine Learning models, but requires complementary expertise to secure their production deployment, long-term operation, and data feeding. The mission is positioned at the interface between the DSI's Data Science and Data Platform teams, with a strong focus on industrialization challenges (build & run).

Mission Objectives

• Industrialize Machine Learning projects developed under Dataiku DSS
• Implement a robust, secure, and sustainable MLOps chain
• Ensure the reliability, performance, and governance of data flows feeding the platform
• Manage the transition from POC to production, and then operational maintenance
• Disseminate MLOps & DataOps best practices to Data teams

Scope of Intervention

• Data Science & AI projects developed under Dataiku DSS
• MLOps chains: model training, deployment, and monitoring
• Connections between Dataiku and the company's data sources
• Management of DEV / TEST / PROD environments

Expected Deliverables

• Industrialized, documented, and maintainable Dataiku MLOps pipelines
• Machine Learning models deployed and monitored in production
• Operational Dataiku connectors to various data sources
• MLOps & DataOps operational documentation
• Recommendations for standardization and continuous improvement

Interactions

• Reporting: Head of the SI Data Platform & BI domain, with operational reporting to the Data Science team during the mission.
• Key partners: IT teams (projects, development, MCO), business departments (commerce, marketing, supply chain, finance), compliance and security teams, as well as Data Scientists, Data Engineers, Data/Cloud Architects, and technical teams (infrastructure, production, security, operations).

Required Skills

• Dataiku DSS (advanced level, MLOps & DataOps)
• Significant experience in industrializing Machine Learning models
• Proficiency in Python and SQL
• Experience with CI/CD tools and Git
• Good understanding of Data architectures
• Experience with Cloud environments (AWS, GCP or Azure)
• Strong sensitivity to security, RUN, and operational challenges