Tom Barber

As decision-making at Virgin Atlantic sped up, our BI stack creaked. Leaders wanted answers in hours and questions increasingly cut across commercial, operational, and customer data. Like many teams, we turned to GenBI to close that gap. Our first attempt didn’t go well. Answers arrived faster, but confidence dropped. Without shared metric definitions, visible lineage, or governance designed in from the start, ambiguity scaled faster than trust. “Is this right?” became the default question again. In this session, we share how we rebuilt GenBI on trusted foundations using Databricks. We’ll show how Unity Catalog and UC Metrics Views anchor AI answers to signed-off metrics, how Databricks Genie enables users to interrogate and explain results, and why explainability became part of the user experience. Attendees will leave with practical lessons for moving to decision-grade GenBI, drawn from real production use cases and delivered as a visual, story-led case study.