Conversation Testing - NLP

Back to NLP Topics

Conversation Testing simulates end-to-end conversational flows to evaluate dialog task execution and support regression testing. Create Test Suites to capture business scenarios and run them to validate app performance.

The framework tracks transition coverage and measures how well the app understands queries and executes dialogs. Flow Health in the Health and Monitoring dashboard summarizes total coverage of dialog flows and conversation test results, helping you identify and fix missing transitions and intent gaps.

Go to: Automation AI > Virtual Assistant > Testing > Regression Testing > Conversation Testing

---

Landing Page

The Conversation Testing page lists all Test Suites in a grid with the following columns:

Column	Description
Test Suite Name	Name of the suite.
Result	Execution status of the latest run.
Duration	Time taken for the latest execution.
Passed / Failed / Not Executed	Test case counts.
Pass %	Percentage of passed test cases.
Tags	Labels added during suite creation for filtering and organization.

Only the results of the latest execution are shown. Earlier run history is not displayed on this page.

Filter options:

• Search and filter any column directly in the grid.
• Filter by intents covered across all test suites — useful for regression after intent definitions change.

Suite actions: Edit, clone, export, or delete a suite using the options menu. Columns can be reordered by dragging.

---

Key Features

Feature	Description
Create a Test Suite	Build test suites to capture conversation scenarios.
Test Editor	Edit and manage individual test cases and steps.
Test Case Assertion	Define assertions to validate app responses.
Test Case Execution Summary	View detailed results after running a test suite.

---