Databricks

Full guide for the Databricks SimpleChat action.

What this action does

Databricks uses the Databricks SQL Statement Execution API for Azure Commercial workspaces. It can execute SQL, list catalogs, list schemas, list tables, and describe a table through a configured SQL Warehouse.

Why and when to use it

Use Databricks when governed analytics data lives in Databricks and users should query it through approved agents instead of direct warehouse tools. It is read-oriented and configured with row/time limits. Do not use it for Snowflake, generic SQL databases, or spreadsheet uploads; use Snowflake, SQL Query, or Tabular Processing instead.

Before you start

Azure Commercial Databricks Workspace URL, not the SQL statement API path.
SQL Warehouse ID.
Optional default Catalog and Schema.
Personal access token, bearer token, service principal, managed identity, or compatible reusable identity.
Agents/actions enabled with enable_semantic_kernel.

Configure the action

Choose Databricks.
Set Workspace URL.
Confirm Cloud is Azure Commercial.
Enter SQL Warehouse ID and optional Default Catalog and Default Schema.
Choose Authentication Method: Personal Access Token, bearer token, Service Principal, or Managed Identity.
Fill Token or Client ID, Client Secret, and Tenant ID when those auth modes are selected, or choose Reusable Identity.
Set Max Rows, Timeout (seconds), and Wait Timeout (seconds).
Use Test Connection.

Example prompts

“Show the top product categories by revenue in the last completed quarter.”
“List tables in the finance schema and describe the invoice table.”
“Run a read-only query that groups daily active users by week for the past 90 days.”

Troubleshooting