databricks-genie

You can query Unity Catalog using natural language by creating a Genie Space and asking questions through the Conversation API. Databricks Genie translates your questions into SQL and returns the generated query along with tabular results.

To migrate Genie Spaces between workspaces with different catalog names, export the serialized_space definition, replace the catalog names within the payload for remapping, and then import the updated definition into the target workspace.

Yes, natural language SQL exploration allows you to query complex datasets without writing SQL manually. Genie Spaces use the Conversation API to translate your natural language questions into SQL queries over your configured Unity Catalog data.

The best way to manage and back up Genie Spaces is by using the management tools to create, update, and export serialized space definitions. These serialized payloads can be stored for backup, cloned for new environments, or imported during workspace migration.

Inspecting table schema before asking questions through the Conversation API is optional but recommended. Reviewing the Unity Catalog schema helps ensure your natural language questions align with the available columns and data structures for accurate SQL generation.

Databricks Genie handles follow-up questions within the same Conversation API session, allowing for conversational data exploration. You can ask sequential questions to refine your query results without needing to re-establish the context of the Genie Space.