Skip to main content

Agent Test Center: How to Create and Run Automated Tests

Written by Miguel from SalesCloser
Updated over 10 months ago

Introducing the Agent Test Center, a tool designed to help you automate testing and continuously improve your AI agents.

You can find it in the left navigation bar of your dashboard. The Test Center allows you to simulate customer calls, evaluate agent responses, and review detailed results — all in one place.

Step 1: Create a New Test

  • Click Create Test in the top right corner of the Agent Test Center dashboard.

  • This will open the Test Wizard.
    Enter a Test Name and select the agent you want to test.

Step 2: Set Up Test Profiles

  • Define the customer profiles your agent will interact with.

  • You can create multiple profiles to simulate different types of customers.

  • Be sure to include important details such as:

    • Customer type (e.g., startup founder, enterprise executive).

    • Key pain points.

    • Possible objections they might raise.

Step 3: Create Test Cases

  • Add test cases that define the key performance behaviours you want to validate.
    For each test case:

    • Write a question (e.g., "Did the agent ask about the customer’s needs?")

    • Set whether a Yes or No answer would be considered a pass.

  • You can add multiple test cases to evaluate different aspects of the agent's performance.

Step 4: Save and Run the Test

  • After completing the setup, click Save Test.

  • You will be directed back to the Test Center dashboard.

  • To run the test:

    • Click Run next to your saved test.

    • Select how many times you want the test to run (up to a maximum of 5).
      Click Run Test to begin.

Step 5: Review Test Results

  • Once the simulation finishes (this may take a few minutes), you can view the results:

    • View Test Profiles and see which customer scenarios were used.
      Review Test Cases to check whether the agent passed or failed specific evaluations.

    • See the Overall Performance Score for quick assessment.

    • Access the Full Transcript to review the entire conversation and spot areas for improvement.

Helpful Tips

  • Use Full Transcripts to pinpoint specific interactions that may need optimization.

  • Running multiple tests can help validate consistency and uncover hidden issues.

  • Regular testing and adjustments can significantly improve your agent’s quality and performance over time.

For any questions, contact us at hello@salescloser.ai or visit our website at https://salescloser.ai/

Did this answer your question?