AB-620: Designing and Building Integrated AI Solutions in Copilot Studio

By Microsoft · Question Mode

QuestionQ14

Test and manage agents

You run multiple evaluation tests for an agent in Copilot Studio before expanding access for users. The tests show the following results:

  • Each evaluation case is reported as meeting or missing the expected response.
  • Some evaluation cases repeatedly fail over several runs.
  • Each run shows the expected response and the response the agent generated.

Provide a conclusion based on the evaluation results.

Select
Evaluation case meeting or missing expected response.
Some evaluation cases fail repeatedly.
Each run displays expected response and agent response.
Explanation

An evaluation case that meets or misses its expected response has a pass-or-fail result. Repeated failures across several runs are consistent outcomes. Comparing the expected response with the agent’s actual response displays the expected and generated responses; it does not alone establish a root cause or authorize deployment.

Learn more

Community Discussion

No comments yet. Be the first to start the discussion!