Sinna Lisa Vange

The presentation shows how Databricks enables non‑technical users in pharma to gain scientific insights by uniting dashboards, chatbots, documents and data. The foundation is data and documents from source systems organized in Unity Catalog and a semantic layer of schemas, tables with definitions, metrics views with business logics, keys linking datasets, SQL functions, Genie and Agent Bricks. Clinical study data is complex, with interpretation depending on context often described in documents. We use a dashboard for context, with visuals and filters conveying the overarching story, while Genie and Agent Bricks offer insights via natural-language questions. Traditional dashboards excel at data storytelling but lack flexibility and tends to grow into overwhelming complexity, by adding context-aware chatbots we address these challenges. Insights can also be embedded into platforms like Teams and SharePoint, ensuring easy access, while having data and calculations centrally governed.