Skip to main content

Azure SQL User Connector Guide

This guide provides step-by-step instructions for creating and configuring an Azure SQL user connector with Abacus.AI. Azure SQL user connector supports user-level permissions, role-based access control (RBAC), and row-level security (RLS)

An Entra admin must set this connector up first

Microsoft blocks ordinary users from consenting to the Abacus.AI app on their own. Before anyone else in your tenant can connect Azure SQL, a Microsoft Entra admin must create the connector once and tick the "Consent on behalf of your organization" checkbox on Microsoft's Permissions requested screen, then click Accept.

This is not the same as the Grant admin consent button in the Entra admin center. Once the admin has ticked the checkbox, every user in that tenant can sign in and create their own connector freely — until then Microsoft stops them with "Need admin approval".

See Microsoft Entra Tenant Consent for the full walkthrough.

Prerequisites​

  • Access to Azure SQL Database
  • Abacus.AI account with appropriate permissions
  • Azure Entra ID credentials

Step-by-Step Instructions​

1. Create Azure SQL Connector​

Create an org level Azure SQL connector using JDBC/ODBC option. Use ODBC 18 driver to connect to Azure SQL server.

Azure SQL connector creation interface

2. Enable RBAC Import​

Turn IMPORT RBAC toggle ON.

RBAC toggle configuration

Note: Please enable the IMPORT RBAC toggle; otherwise, users will use the ODBC connection and might gain access to more data than expected.

3. Configure Connection Details​

Fill the details and click on CREATE.

Connection details form

4. Database User Setup​

Create in your database and whitelist the IPs.

CREATE USER [abacus_azure_sql] FROM EXTERNAL PROVIDER;

5. Verify Connector Status​

Click on Verify Now and make sure that connector is in active state.

6. Navigate to Projects​

After creating the connector, please navigate to the projects page by clicking on the Abacus.AI logo.

Navigation to projects page

7. Create New Project​

Create a new project.

8. Select ChatLLM Option​

Select GenAI --> Custom Chatbot option.

9. Enter Project Name​

Enter the name of the project.

10. Skip to Dashboard​

Select skip to project dashboard.

11. Access Model Training​

Click on the Model option in the Left tool bar and select Train Model on the top right corner of the page.

Data source configuration

12. Select External Service​

From the dropdown in Structured data source, select External Databases.

Structured Data Source Options

13. Select Connector and Tables​

Select the connector you create in step 1 from the dropdown and add the tables that can be used for testing.

Model training initiation

14. Train Model​

Click on Train Model.

15. Access Trained Model​

Once training is complete. Click on models and select your model.

Trained model access

16. Create Deployment​

Click on create a new deployment.

Deployment creation interface

17. Select Deployment Type​

Select Offline Batch + Realtime. Click on Next.

18. Configure Deployment Name​

Enter the name that you want your end users to see and click on deploy.

19. Monitor Deployment Status​

Go back to your model by clicking on model and then your model name. Once the deployment is in active state. Click on deployment name.

20. Access Prediction Dashboard​

Click on the prediction dash option under Deployment.

21. Test the Chat Bot​

Click on Go to Abacus.AI chat to test your bot.

22. Authentication and Testing​

Enter your question/prompt. Each user will be asked to login to Azure SQL once. Login using your Entra ID user credentials.

Important Notes​

  • Ensure that your Azure SQL database is properly configured with the necessary permissions
  • The ODBC 18 driver is required for optimal compatibility
  • Each user will need to authenticate with their Entra ID credentials when first accessing the system
  • Make sure to whitelist the appropriate IP addresses for your Azure SQL database
  • The connector must be in an active state before proceeding with model training

Troubleshooting​

If you encounter issues during the setup process:

  1. Verify that the Azure SQL connector is properly configured and active
  2. Check that the database user has been created correctly
  3. Ensure that IP addresses are properly whitelisted
  4. Confirm that Entra ID authentication is working properly

This completes the setup process for the Azure SQL RBAC user connector with Abacus.AI.