From Data Pipelines to Agentic Workflows

Data platforms are more powerful than ever, yet preparing data still depends heavily on scarce engineering expertise. Agentic workflows offer a new approach: AI-supported systems that inspect data, identify issues, suggest transformations and guide users towards analytics-ready datasets. In this session, we explore how agentic workflows change analytics engineering and walk through a practical process for designing one: selecting the right workflow, mapping decisions, defining agent responsibilities, adding tools and context, and implementing human oversight and governance