Summary
In this module, you explored how Azure Databricks Genie makes your organization's data accessible inside Microsoft Teams and Microsoft 365 Copilot—without requiring any technical skills or SQL knowledge.
What you learned
- Genie surfaces: Azure Databricks Genie is available through Genie Agents, Genie One, the Genie mobile app, the Databricks Genie app in Microsoft Teams, and Genie on Microsoft 365 Copilot. Choose the surface that best fits where you're working.
- Using Genie in Teams: You can access Genie through a direct message or by mentioning
@Databricks Geniein a channel. The Teams app supports natural-language questions, multi-turn threads, source citations, suggested follow-up prompts, and feedback, but it doesn't display charts or visualizations. - Using Genie in Microsoft 365 Copilot: You can open a persistent Genie conversation via More agents or invoke Genie with
@Databricksin a new chat. The same feature limitations as Teams apply, including no chart support. - Asking effective questions with SPARK: The SPARK framework (Scope, Period, Aggregation, Result, Key metric) helps you build specific questions that reduce ambiguity and produce reliable answers. Multi-turn follow-up questions build on earlier context within the same thread.
- Interpreting and verifying answers: The inspect-verify-refine workflow—checking filters, reviewing source citations, spotting anomalies, and following up on uncertain answers—helps you translate a Genie response into a decision with confidence.
- Using AI-generated answers responsibly: Genie answers factual data questions, not interpretive or advisory ones. It can generate inaccurate responses; your accountability for decisions based on those answers doesn't change. Consider visibility when sharing answers in channels, and escalate to a domain expert or your data team for high-impact or anomalous findings.