Skip to main content

Can enterprise search answer questions from dashboards and data, not just documents?

Summary

  • Yes. On Databricks, AI/BI Genie answers plain-language questions from your governed structured data and from your dashboards, not only from documents.
  • Genie translates a natural-language question into a SQL query against annotated tables and returns the answer as a text summary, a results table, and a visualization, along with the query it ran.
  • Published AI/BI dashboards include an Ask Genie button, so users can ask follow-up questions about the dashboard's data beyond the predefined charts.
  • Genie runs inside Unity Catalog governance, so answers respect the access controls you already have.
  • The Genie Conversation APIs let you embed the same natural-language answering into custom apps and collaboration tools, so people can query data where they already work.

Can enterprise search answer questions from dashboards and data, not just documents?

Yes. Many search experiences are built for documents and other unstructured content, but business questions are often answered by structured data, the tables and dashboards behind day-to-day operations. On Databricks, AI/BI Genie is the natural-language interface to that governed data: a user asks a question in plain language and Genie returns the answer as a text summary, a table, and a visualization, drawn directly from your data and dashboards rather than from documents alone.

Why Databricks Genie answers questions from data and dashboards

  • Natural language over governed tables. Genie is configured by domain experts with datasets, sample queries, and text guidelines. It selects relevant names and descriptions from annotated tables and columns to convert a natural-language question into a SQL query, then responds with the query and a results table, or asks a clarifying follow-up when needed. See Genie concepts.
  • Ask questions directly from a dashboard. Published AI/BI dashboards include an Ask Genie button. When you publish a dashboard, Databricks automatically generates a companion Genie Agent based on that dashboard's datasets, so viewers can dig deeper, for example asking for sales by region for a particular product line, without leaving the dashboard. See what's new in AI/BI dashboards.
  • Governed by Unity Catalog. Genie operates on Unity Catalog-governed tables, so answers stay secure and consistent and each user sees only the data they are permitted to access. See AI/BI concepts.
  • Answers where people work. The Genie Conversation APIs let you embed the same natural-language answering into custom applications, Databricks Apps, and collaboration tools, so a team can ask questions about data from the surfaces they already use while governance is maintained.
  • Transparent and honest about uncertainty. Genie returns the SQL it generated so answers can be reviewed, keeps conversation context so questions can be refined, and asks clarifying questions instead of returning an incorrect result when a request is ambiguous.

Getting started

FAQs

Can it answer questions from structured data, not just documents?

Yes. Genie translates a plain-language question into a SQL query against your governed tables and returns a text summary, a results table, and a visualization, so it answers from data rather than only from documents.

Can users ask questions about a dashboard?

Yes. Published AI/BI dashboards include an Ask Genie button, and publishing a dashboard automatically creates a companion Genie Agent from its datasets, so users can ask follow-up questions beyond the predefined charts.

Are the answers governed?

Yes. Genie runs on Unity Catalog-governed data, so every answer respects the access controls already in place and each user sees only the data they are authorized to access.

Can we embed this into our own applications?

Yes. The Genie Conversation APIs let you embed natural-language question answering into custom apps, Databricks Apps, and collaboration tools while maintaining governance.

The information provided herein is for general informational purposes only and may not reflect the most current product capabilities or configurations.