Skip to main content
Delphina connects to Snowflake using a dedicated user and role. The role needs SELECT access to the data you want the agent to query, and — optionally — access to query history and Cortex Analyst objects so Delphina can learn from existing usage and semantics.
We recommend the warehouse provided be sized Small or larger so Delphina can execute queries with acceptable performance.

Connecting Snowflake

Step 1: Gather connection details

You’ll need the following from your Snowflake account:

Step 2: Create the user, role, and grant data access

Run the following DDL in Snowflake. It creates a dedicated user and role and grants SELECT access to every object in a database. Prune the grants down to a smaller set of schemas/tables if you’d rather scope Delphina narrowly.

Step 3: Grant access to query history

Delphina uses query histories from trusted users to prioritize tables and learn how your data is used. Choose one of the two patterns below.
Give the role read access to the shared SNOWFLAKE database, which exposes ACCOUNT_USAGE.QUERY_HISTORY.

Step 4: (Optional) Grant access to Cortex Analyst objects

If you use Snowflake Cortex Analyst and want Delphina to reuse your existing semantic definitions, grant the following additional permissions:

Step 5: Create the connection in Delphina

  1. Navigate to analytics.delphina.ai.
  2. Click your name in the bottom-left, then Org Admin > Warehouse Connections.
  3. Click Add Connection on the target workspace.
  4. Set the Warehouse Type to Snowflake.
  5. Fill in the account identifier, warehouse, database, role, and user credentials from the steps above.
  6. Click Create Connection, then Test Connection to verify Delphina can reach your warehouse, list tables, and access query history.

Troubleshooting