Create a new assistant test
Creates a comprehensive test configuration for evaluating AI assistant performance
Authorizations
Bearer authentication header of the form Bearer <token>, where <token> is your auth token.
Headers
Optional opaque, unquoted key for safely retrying the same logical request. Keys must contain 1 to 255 letters, numbers, hyphens, or underscores. Generate a unique UUID v4 for each operation and reuse it only when retrying that operation with the same request. Invalid headers—including duplicate, empty, malformed, or overlong values—return 400 with error code 10015. A request already in progress with the same key returns 409; reusing the key with a different request returns 422. Only successful responses are replayed, for up to 24 hours. Do not include sensitive data in the key.
1 - 255^[A-Za-z0-9_-]{1,255}$Body
Request model for creating a new assistant test.
This model defines all the required and optional parameters needed to create a comprehensive test for an AI assistant, including test configuration, evaluation criteria, and scheduling options.
A descriptive name for the assistant test. This will be used to identify the test in the UI and reports.
1 - 255"Customer Support Bot Test"
The target destination for the test conversation. Format depends on the channel: phone number for SMS/voice, webhook URL for web chat, etc.
1"+15551234567"
Detailed instructions that define the test scenario and what the assistant should accomplish. This guides the test execution and evaluation.
1 - 5000"Act as a frustrated customer who received a damaged product. Ask for a refund and escalate if not satisfied with the initial response."
Evaluation criteria used to assess the assistant's performance. Each rubric item contains a name and specific criteria for evaluation.
1Optional detailed description of what this test evaluates and its purpose. Helps team members understand the test's objectives.
1000The communication channel through which the test will be conducted. Determines how the assistant will receive and respond to test messages.
phone_call, web_call, sms_chat, web_chat "web_chat"
Maximum duration in seconds that the test conversation should run before timing out. If not specified, uses system default timeout.
1 <= x <= 3600Optional test suite name to group related tests together. Useful for organizing tests by feature, team, or release cycle.
100Response
Returns the created test configuration with assigned test ID
Response model containing complete assistant test information.
Returns all test configuration details including evaluation criteria, scheduling, and metadata. Used when retrieving individual tests or after creating/updating tests.
Unique identifier for the assistant test.
"123e4567-e89b-12d3-a456-426614174000"
Human-readable name of the test.
"Customer Support Bot Test"
Communication channel used for test execution.
phone_call, web_call, sms_chat, web_chat "web_chat"
Evaluation criteria used to assess test performance.
Timestamp when the test was created.
"2024-01-16T09:00:00Z"
Detailed description of the test's purpose and scope.
Target destination for test conversations.
Maximum allowed duration for test execution in seconds.
Test suite grouping for organizational purposes.
Detailed test scenario instructions and objectives.