What AI can search my emails, docs, and databases at once?

Summary

• Databricks Genie is a conversational AI that lets you ask questions in natural language across your databases and documents at once — no SQL required.
• For structured data, Genie translates your question into governed SQL run against the tables, dashboards, and queries registered in Unity Catalog.
• For unstructured content, built-in connectors and Model Context Protocol (MCP) support let Genie pull context from enterprise systems and document repositories such as Google Drive and SharePoint.
• The ai_search() function retrieves and reranks results across multiple knowledge sources and synthesizes a grounded, natural-language answer.
• Every interaction runs inside Unity Catalog governance: users see only data they are authorized for, queries are read-only, and each request is logged for audit.

What AI can search my emails, docs, and databases at once?

Databricks Genie is the AI interface for asking questions in natural language across your enterprise data — both the structured data in your databases and the unstructured content in your document repositories — from one conversational experience. Instead of writing SQL or hopping between tools, you type a question and Genie determines where the answer lives, retrieves it under your existing permissions, and responds in plain language. It combines database access with document search, so a single question can draw on tables, dashboards, and business documents together.

Why Databricks Genie

• Natural-language questions over your databases. Genie translates your question into governed SQL executed against the tables, dashboards, queries, and notebooks registered in Unity Catalog, so business users get answers without writing code.
• Access to documents and enterprise systems. Built-in connectors and Model Context Protocol (MCP) support let Genie pull context from enterprise systems and document repositories such as Google Drive and SharePoint, bringing knowledge stored outside traditional databases into the same conversation.
• Search and synthesis across sources. The ai_search() function retrieves information from one or more AI Search knowledge sources: given a natural-language query, it generates optimized search queries, retrieves and deduplicates results across sources, reranks them by relevance, and synthesizes a grounded, natural-language answer.
• A unified chat experience. Chat in Genie provides a full-screen interface for asking data questions in natural language, searching across your existing dashboards, queries, and Genie agents while also connecting to external data sources.
• Business context for accurate answers. A semantic layer — Genie's ontology — captures business terms, metric definitions, and table relationships, so answers are grounded in how your organization actually defines its data.
• Governed by Unity Catalog. Every interaction runs within Unity Catalog access policies: users see only the data they are authorized to access, queries are read-only, connections are governed through the Unity Catalog AI Gateway, and each interaction is logged for audit.
• Available everywhere your teams work. Through the Genie Conversation APIs, Genie can be embedded into Databricks Apps, Slack, Teams, SharePoint, and custom applications, so natural-language search reaches users on any surface.

Getting started

• Read The next generation of Databricks Genie for built-in connectors and MCP support.
• Learn the fundamentals in Genie agents concepts, including what data you can use.
• Explore the ai_search function for retrieval across knowledge sources.
• Embed Genie with the Genie Conversation APIs.

FAQs

What is Databricks Genie?

Genie is a conversational AI interface that lets business users ask questions in natural language and get answers from enterprise data. It translates questions into governed SQL for structured data and uses connectors and search to reach unstructured documents.

Can Genie search both databases and documents?

Yes. Genie queries structured data registered in Unity Catalog and, through built-in connectors and MCP support, pulls context from document repositories and enterprise systems such as Google Drive and SharePoint.

How does Genie keep answers secure?

All Genie interactions run within Unity Catalog access policies. Users see only data they are authorized to access, queries are read-only, connections are governed through the Unity Catalog AI Gateway, and every interaction is logged for audit.

Where can I use Genie?

Genie is available in the Databricks workspace and can be embedded into Databricks Apps, Slack, Teams, SharePoint, and custom applications through the Genie Conversation APIs.