Skip to content

raia Copilot · Section

Simulations & Testing

Run scripted conversations against any agent version.

Simulations run scripted conversations against any agent version — production, staging, or a draft — so you can prove behavior before customers see it. Each turn carries pass/fail criteria, which makes regressions visible instead of anecdotal.

In this section

  • Author a simulation as a sequence of user turns with expected outcomes.
  • Run a suite against multiple agent versions and compare results side by side.
  • Interpret failures using the same retrieval trace you get in Admin Mode.
  • Gate releases on a passing suite before flipping an agent live.
01

What's inside

  1. 01

    Scenario

    A named script: the persona of the simulated user, the channel, and the ordered list of user turns.

  2. 02

    Assertions

    Per-turn criteria — must contain, must not contain, must call a named function, must escalate, must stay under a latency or cost budget.

  3. 03

    Run

    Execute against a chosen agent version. Every turn records the reply, the skills called, the retrieval sources, and the assertion result.

  4. 04

    Report

    Pass rate per scenario and per assertion, a diff against the previous run, and a link into each failing conversation.

02

At a glance

SuiteWhat it coversRun it when
SmokeFive to ten core questions the agent must always answer correctly.Every change, before publishing
KnowledgeQuestions that map to specific documents in the vector store.After any Train tab change
ActionsFlows that must trigger a Function, Webhook, or Calendar call.After changing a skill or schema
GuardrailsPrompts that must be refused, redacted, or escalated.Before a compliance review
RegressionEvery scenario that previously failed in production.Weekly, and before major releases

Tip

Turn every production incident into a scenario. A suite built from real failures is worth far more than one written from imagination.

03

Related