Extracting Business Information from Cases

This lesson explains how to set up Business Information Extraction scenarios using Generative AI within SAP Service Cloud Version 2.

It guides administrators through the essential steps to activate and configure extraction scenarios, map relevant Case fields, and connect business documents to Case types. These setups enable the automatic extraction of important data from Case content and attachments, making business processes more efficient.

A potential use case for the generative AI-based information extraction feature could be to pull business details from the subject or description (such as the Product ID, Product Name, Quantity, Serial ID, etc.) of a case and add that data to the business documents table (like the Service Order).

The first activity you, as an Administrator, must perform after addressing the commercial and technical prerequisites is to activate and configure the business information extraction at the following path: Settings→All Settings→Machine Learning→Generative AI→Business Information Extraction.

Once the Business Information Extraction feature is enabled, follow these steps to complete the configuration:

1. Select Elements from the left pane and click the plus icon.
2. Enter a name, label, and description for your element.

   The description helps the LLM identify the content in attached documents. Describe the type of content and, if relevant, where this content appears on a form.

3. Enter a Short Sample. This brief sample helps the system identify your elements. Providing at least one short sample is required.
4. Enter sample values, separated by commas, then click Save. These are the values of your elements.
5. Select Classes from the left pane and click the plus icon.
6. Enter a name, label, and description for your class.

   The description helps the LLM to classify content from your documents.

7. Enter Short Sample, Value, and click Save. At least one short sample and a value are required for the system to identify class information.
8. Select Scenario from the left pane and click the plus icon.
9. Enter a scenario name and a detailed description of the scenario. The scenario description should provide the context and role in which Generative AI is expected to act, helping to extract elements and classify content. For example, the scenario name could be 'Extract Product Name from Case,' and the description might be 'Extract Product Name from a newly created case detail.'
10. Select and add elements in the Element Selection section and select classes in the Class Selection section.
11. Perform a test extraction by entering sample text in the text box and clicking Test. Once the information is extracted, you will see the extracted elements and class derivations with confidence percentage values.

You have now successfully set up the Business Information Extraction capability on your system.

After creating your first extraction Scenario, you can now move on to the next step, which involves adding this Scenario to the Case Type.

You can add the Business Information Extraction Scenario to an existing Case type by navigating to Settings→All Settings→Case Types . Select the Case Type where you want to include the Scenario and go to the Case Detail view.

Inside the Case Detail settings page, go to the "Additional Features" tab and select the extraction scenario from the drop-down list.

Activate the new version of the Case type.

You have now successfully set up a Business Information Extraction Scenario for your Case type in SAP Service Cloud Version 2.

In this lesson, we:

• Explained how Business Information Extraction can be activated using Generative AI units.
• Described a potential Use Case where an extraction scenario can be implemented.
• Demonstrated how to configure an extraction scenario and how to consume it from a Case Type.
• Assigned the extraction scenario to an existing Case Type and Activated it.