Testing LLM topics

  • Release version: Australia
  • Updated April 6, 2026
  • 2 minutes to read
    • Summarize
    Summarized using AI
    This content was generated using new OpenAI-powered functionality. Results are provided on an as is basis and are not guaranteed to be accurate or complete.

    Summary of Testing LLM topics

    This feature enables ServiceNow customers to preview, test, and debug topics that utilize large language models (LLMs) directly within the Assistant Designer chat window. It supports testing in the Service Portal and in third-party messaging applications like Microsoft Teams and Slack, allowing you to verify how your Virtual Agent topics perform across different deployment environments.

    Show full answer Show less

    To use this functionality effectively, you must have the Now Assist Topics skill activated, which provides access to the updated Assistant Designer Asset library interface.

    How to Test LLM Topics

    • Navigate to All > Conversational Interfaces > Assistant Designer and select the Asset library tab.
    • Choose the LLM option and select the topic you want to test.
    • Click Test to open the chat test window in the Now Assist chat widget.
    • If a topic is linked to multiple LLM assistants, use the Test skill discovery checkbox and the Assistant drop-down to pick the assistant to test.
    • If Microsoft Teams or Slack is configured, test options for those channels appear under the Test button drop-down.

    Issue Identification and Feedback

    The system visually flags incomplete or problematic topics with badges indicating the number and location of issues. Attempting to test an incomplete topic opens an Issues window listing detailed descriptions and links to each problem area.

    During testing, you can provide direct feedback on each LLM-generated utterance by using the thumbs-up or thumbs-down icons, helping to refine the topic’s performance.

    Test Information and Next Steps

    The chat test window includes tabs like Analyze test phrases, Variables, Context, and Logs, which provide insights into the topic’s behavior and performance during testing.

    After testing, close the test window and use the gathered data to improve your topic. For example, low or uncertain confidence ratings on the Analysis tab suggest updating topic descriptions or instructions within LLM user input nodes.

    Preview, test, and debug topics that use large language models (LLMs) in the Assistant Designer chat window.

    Note:
    An updated Assistant Designer Asset library user interface is available when you install Now Assist in Virtual Agent and turn on the Now Assist Topics skill. This content assumes that you have activated this skill and can see the list view. If this skill is not activated, you see the legacy UI and topics page. For more information, see Virtual Agent Designer legacy topics page.

    About testing LLM topics

    You can test your topic's functions by running your conversation in a chat test window as you work. The web (Service Portal) chat client is the default test window.

    If you're using the Virtual Agent integrations with third-party messaging apps, elements in your conversation might appear differently in third-party messaging applications. Test your conversations in any third-party applications where you want to deploy Virtual Agent.

    Note:
    If the Now Assist panel, Microsoft Teams application, or Slack application is configured for your environment, preview options for those channels are displayed in the Test button list. Select Preview in Now Assist panel or Preview in Microsoft Teams in the list to test your topic in those environments.

    Starting a test

    You can find LLM testing option on the Virtual Agent Designer canvas.

    1. Navigate to All > Conversational Interfaces > Assistant Designer.
    2. Select the Asset library tab.
    3. Select the LLM option.
    4. Select the topic that you want to test.
    5. Select Test.

      Assistant Designer Asset library list and topic canvas views. Test options on both are highlighted.

    The chat test window opens in the Now Assist in Virtual Agent chat widget.

    If the topic is associated with multiple LLM assistants, select the Test skill discovery check box, then open the Assistant drop-down list to select which assistant you want to test. For more information about LLM assistants, see LLM assistants.
    Note:
    The Test in enhanced chat option is available only from the Asset library list view. The Test button in the topic canvas adds options in a drop-down box if you configure Microsoft Teams or Slack for Virtual Agent.

    Add assistants in the Properties tab, followed by the Test skill discovery check box in the topic testing window, to select an assistant in the drop-down window.

    Note:
    If your topic is missing any necessary information, an incomplete badge appears in the corner of the flow diagram pane. A warning badge also appears next to each node. The incomplete badge lists the total number of issues, while the local warning badges show how many issues are found in each node. If you select Test when there are issues, an Issues window opens with the number of issues present along with the details of each issue. These details include a full description and a hyperlink to each incomplete item.

    Feedback

    You can provide feedback to each utterance made by the LLM by selecting from the like thumbs-up icon or dislike thumbs-down icon options that appear when you hover over an utterance. All LLM user inputs except the Input Collector have these feedback options.

    LLM feedback options.

    If you run tests from the Topics page, the test window shows only the Analyze test phrases, Variables, Context (available by default, with no Include topic discovery option), and Logs tabs.

    Next steps

    When you're done testing your topic, close the test chat window. If necessary, use the test information to adjust your topic to perform more accurately. For example, the results on the Analysis tab may return low scores or Unsure or Unknown confidence ratings. Improve scores by updating the topic description or instructions in the LLM user input nodes.