Connect your own Microsoft Azure AI Document Intelligence model

If your organization’s security protocols or data privacy policies require the use of your own subscription to Microsoft Azure AI Document Intelligence models, it is necessary for you to connect your Microsoft Azure AI Document Intelligence account with Document Automation.

Prerequisites

  • Ensure that you have your Control Room authorization token. See Authentication API (V2).
  • Ensure that you have the following information:
  • If you are using an On-Premises Control Room, ensure that you add the following API URLs to your trusted list:
    • Add the Azure AI Document Intelligence services API URL to the trusted list at the Bot Runner end.
    • Add the SAS (blob) URI to the trusted list at the Control Room end.
    • Add the Azure AI Document Intelligence services API URL to the trusted list at the Control Room end.
    • Add the https://fotts.azureedge.net/ URL (required for specific document type rendering) to the trusted list at the Control Room end.

To connect to Azure AI Document Intelligence using your own subscription, you must send an API request (as detailed on this page). After the request is successful, you will be able to view your Azure AI Document Intelligence models when you create a learning instance for the Standard Forms document type and click Browse to select a model. See Create a learning instance for standard forms.

Note: You will not be able to see the Create Model option on the learning instances page to create models when you use your own license (BYOL). See Build and train a custom extraction model.

Procedure

Run the following API to connect your Azure AI Document Intelligence model for Standard Forms:

Endpoints:

  • Cloud: PUT <Cloud Host Url>/cognitive/v3/formrecognizer/config
  • On-Premises: PUT <On-Premises Machine ip_address or localhost:port>/cognitive/v3/formrecognizer/config

Authentication: Headers: X-Authorization - <Document Automation admin user authentication token>

Request body schema: application/json

{    
"formRecognizerApi": "https://<Cognitive Service Endpoint>",    
"formRecognizerKey": "<Cognitive Service apiKey>",    
"blobStorageSas": "https://<Blob storage URI>" 
}
Parameter Format Description
formRecognizerApi (required) String Azure AI Document Intelligence cognitive service endpoint URL.
formRecognizerKey (required) String Azure AI Document Intelligence cognitive service API key.
blobStorageSas (required) String Azure Blob Storage URI.

After your request is submitted successfully, you will see the status code as 200 OK. You can then choose your Azure AI Document Intelligence models when creating a learning instance using the Standard Forms provider. See Create a learning instance for standard forms.