Skip to main content
Use Conductor to investigate unsuccessful calls and chats, then test improvements with simulations against your agent. Test a proposed fix or an edge case, such as a wrong-number call, before using the updated agent with customers.

​
Start from real conversations

Use Call and chat history to investigate unsuccessful sessions before deciding what to test. Turn the problem into a repeatable scenario, then check whether the proposed agent change resolves it.

​
Test and refine edge cases

Edge cases are the unusual situations that trip agents up — an angry caller, a wrong number, someone who changes their mind. Ask Conductor to help you handle them:
  • “What edge cases should I worry about for this agent?”
  • “Make sure the agent handles a caller who refuses to give their name.”

​
Run simulations

Conductor can run simulation tests that play out a scenario against your agent and show how it responds. This lets you check changes without making real phone calls. See Testing with Conductor for creating test cases from calls, configuring mocks, and running a saved suite.
  • “Run a simulation where the caller wants to reschedule an appointment.”
  • Attach a saved test case and say “Run this and tell me if it passes.”
When a simulation finishes, Conductor shows the result so you can see how your agent behaved and decide what to refine next.
Conductor Simulation Tests result showing Wrong number - caller not the patient passed, with one test completed and a reply explaining that the template agent apologized and ended the call.

A wrong-number simulation passes for Patient Screening (from template).

​
Manage test cases and runs

Ask Conductor to list or inspect saved test cases, update their scenarios or success criteria, or delete cases you no longer need. You can also ask it to run a batch, report the pass rate, and inspect individual run transcripts. Use Add context (+) → Test Cases or Testing History on the agent page to attach the case or run you mean. See Testing with Conductor for personas, dynamic variables, function mocks, and batch workflows. Simulation runs have their own testing charges, separate from Conductor messages.

​
A good loop to follow

1

Make a change

Ask Conductor to build or adjust something, then accept the proposal.
2

Test it

Run a simulation or review a relevant call or chat to see how the change behaves.
3

Refine

Tell Conductor what to improve, and repeat until you’re happy.
See the testing guides for the LLM Playground, simulation testing, and audio testing.