Create an AI connection for Snowflake

Create an AI connection for Snowflake in AI Control Tower using the  AI Service Graph Connector for Snowflake (Version 2.0.5).

Before you begin

Role required: sn_ai_disc.discovery_admin and sn_cmdb_int_util.sgc_admin

Procedure

1. Navigate to Workspaces > AI Control Tower > Configurations > AI connections.
2. Click Add.
3. Select AI Connector for Snowflake from the available connectors.
4. Click Create connection.
5. Review setup instructions page displays the prerequisites for setting up Snowflake Keypair connection.
   Note:

   Verify to follow all the prerequisite steps.
6. Verify to follow all the prerequisite steps.
7. Select Next.
8. Enter the connection details:

Table 1. Fields of connection details

Connection name	A unique identifier for this connection. For example: SGC_Snowflake_KeyPair_Connection
Connection URL	Your Snowflake account URL in account_locator format. For example: https://<account_locator>.snowflakecomputing.com
Snowflake username	The service account name created in Snowflake.
Public Key Fingerprint	The SHA256 fingerprint of your public key from Snowflake.
Certificate (JKS)	The X509 certificate (sys_certificate record) containing your Java Keystore with private key.
Certificate Password	The password for your keystore.
Key Alias	The alias for your private key within the keystore..

Note:

For detailed instructions on configuring JWT key-pair authentication, see the Configuring Keystore for Snowflake Keypair authentication [KB2834688] article in the Now Support Knowledge Base.

9. Enter the details on Create and test connection:
   1. Click Create and Test Connection to validate connectivity.
   2. If the test succeeds, the connection is created and you can proceed to schedule imports.
   3. If validation fails, verify your Snowflake credentials, network policies, and key-pair configuration.
10. Configure import schedule:
    Note:

    After creating a connection, enable and configure scheduled imports.
11. Enable Scheduled Jobs.
    Note:

    Two parent scheduled jobs are created by default but are inactive:

    • Discovery (Agents) – Discovers AI agents and assets.
    • Execution (Usage) – Collects usage and observability data.
12. To enable imports, select the Active check box for both scheduled jobs.
    Note:

    Configure the run frequency for each scheduled job based on your organizational needs. For example, run discovery daily to capture new assets and run execution hourly to monitor usage.
13. Execute on Demand.
    1. To import data immediately without waiting for the scheduled time, click Execute Now next to the scheduled job.
    2. Once configured, scheduled jobs run automatically according to their schedule.
    3. Select Continue.
14. Select the Confirm connection setup activity to verify whether the connection was configured.

Result

Click View all connections to view the newly created connection.

The AI connection for Snowflake is created and configured.