Creating use cases for advanced AI agents

What's my plan?

Add-on

AI agents - Advanced

Use cases are the mechanism by which zero-training AI agents and agentic AI agents (EAP) understand what a customer is asking about and connect them with the right dialogue or procedure (EAP).

Use cases leverage generative AI to power AI agents without the time-consuming activity of training expressions. This approach simplifies setup and maintenance while enhancing the AI agent's understanding and performance.

This article contains the following topics:

• Creating use cases manually
• Creating use cases from suggestions

Related articles:

• Managing use cases for advanced AI agents
• Best practices for creating use cases for advanced AI agents

Creating use cases manually

An AI agent can start supporting customers with as little as one use case.

Tip: Before creating your first use case, see Best practices for creating use cases for advanced AI agents.

As an alternative to creating use cases manually, you can create them based on AI-powered suggestions.

To create a use case

1. In the top-right corner, use the AI agent drop-down field to select the AI agent you want to configure use cases for.
2. In the main menu on the left, select Content > Use cases.
3. Click Add Use Case.
4. Fill out the following fields:
   • Name: Enter a short, descriptive name for the use case. For example, “Order status.”
   • Visitor request reason: Enter a short description that identifies why a customer is reaching out. For example, “Customer is asking for the status of their order.”
5. Click Confirm.

   

Creating use cases from suggestions

As an alternative to creating use cases manually, you can create them based on AI-powered suggestions. The use case suggestions report analyzes your historical conversations to suggest use cases specifically tailored to the needs of your end users. The suggestions include automatically generated use case names, descriptions, and categories, saving you time and effort.

Use case suggestions are a great tool for getting your AI agent implemented quickly as well as regularly improving it. Instead of requiring admins to review historical conversations for potential use cases to create manually, suggestions appear automatically, and admins can quickly review and create them.

To run the use case suggestions report

1. In the top-right corner, use the AI agent drop-down field to select the AI agent you want to configure use cases for.
2. In the main menu on the left, select Content > Use cases.
3. Click the Suggestions tab.
4. Click Generate suggestions.
5. In the Generate use case suggestions window, select a Source:
   • This AI agent: The historical data from the AI agent you selected above.
   • Zendesk Support: Historical email tickets from your Zendesk account. Messaging tickets are not included.
6. In the Conversation sample field, select the number of messages that the system should analyze.

   The options are 250, 500, 750, and 1000.

7. In the Start date field, select the date the system should start looking for meaningful messages.

   

8. (Optional) If you selected Zendesk Support as your source, click Advanced settings to show the following additional options:
   • Filter by departments: Select a specific department from the drop-down list from which conversations should be selected for analysis.
   • Import tags: Select this option to import Zendesk tags as labels.
9. Click Generate suggestions.

   The use case suggestions report begins to analyze your historical conversations. After the analysis is complete, use case suggestion information appears in the following columns:

   • Freq. (Frequency): Shows the estimated percentage of analyzed conversations that a suggested use case applies to. The higher the frequency, the more important it is to create a use case that covers this topic.
   • Use case: Shows a name and description for a suggested use case.
   • Category: Shows a category for a suggested use case.
   • Conversations: Shows the number of conversations that a suggested use case applies to. Click the conversation icon to open a Details panel with more information about the relevant conversations.
     Note: The use case suggestion report is incremental, meaning each time you'd run the report, new use case suggestions are added to already-existing suggestions.

   Additionally, a banner at the top of the tab shows you a summary of what information is covered by the use case suggestions report, including the following details:

   • Conversations: The number of conversations the report analyzed.
   • Suggestions: The number of use case suggestions that were generated.
   • Content coverage: An estimate of the potential increase in use case coverage for your customer conversations.
   • Source: The source you selected when you ran the report.
   • Last updated: The last time the report was updated.

     

10. For each suggested use case, accept it as is, edit it before accepting it, or reject it altogether:
    • To accept a suggested use case, click the checkmark.

      The use case is accepted and added to the Use cases tab.

    • To edit a suggested use case before accepting it, click the use case name to open it in a new window:
      • Edit any of the details as needed. For help, see Creating use cases manually.
      • Click Save & Accept. The use case is accepted and added to the Use cases tab.
    • To reject a suggested use case, click the X.

      Rejected use cases aren’t suggested again the next time you run the use case suggestions report.