› Integration › Recognition providers › Configure a recognition provider using Azure AI Document Intelligence Service

Configure a recognition provider using Azure AI Document Intelligence Service

Use the Azure AI Document Intelligence Service for extracting insights and information from unstructured document data using advanced machine learning and natural language processing. This automates document processing, extracts key information, and enables better decision-making.

To use the Azure AI Document Intelligence Service in a TotalAgility On-Premise environment, ensure the following:

  • Configure a dedicated Microsoft Azure account.

  • Download the mcr.microsoft.com/azure-cognitive-services/form-recognizer/read-3.1 Docker container and follow the instructions provided by Microsoft to install the container.

  • Add relevant connection information.

  • Since the Docker container provides the API key for the Azure AI Document Intelligence Service, you must define the endpoint URL and provide any text for the API key in the TotalAgility Advanced Studio. The API key field cannot be empty.

  1. Navigate to Integration > Recognition providers.

    The Recognition providers page appears.

  2. Click New.

    The New recognition provider configuration dialog box is displayed.

  3. On the Type list, select the provider type as Azure AI Document Intelligence Service.
  4. By default, the ID and Display name are displayed the same as the ID of the provider. The system identifies the OCR provider with this ID. However, you can change the ID and display name as needed.

    You cannot edit the ID once the configuration is saved.

  5. In the Endpoint box, enter the endpoint URL that is used to access the Microsoft Azure AI Document Intelligence Service.
  6. For Authentication, select either option:

    • API key: Enter the API key provided by the Docker container that is used to access Microsoft Azure AI Document Intelligence for recognizing OCR documents.

    • OAuth: On the OAuth server list, select the configured OAuth server to access the Microsoft Azure AI Document Intelligence Service for recognizing OCR documents.

  7. On the Model list, select either OCR option:

    • OCR only: Use this option for powerful OCR engines focused on providing the best text output

    • OCR and layout: Use this option to add structural understanding to the document. This option uses the read model as its foundation and is suitable for more comprehensive data extraction.

  8. Click Save.
See also: