> ## Documentation Index
> Fetch the complete documentation index at: https://docs.callkaro.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Run a Manual RCA

> Analyse a call set you choose yourself, on demand

The daily run covers yesterday, across the whole configured scope. When you need an answer right now — or about a specific set of calls — use **Run Manual RCA**.

<img src="https://mintcdn.com/callkaroai/YnMJfeuar7Le0tPd/images/kpi-manager/manual-rca-select-calls.png?fit=max&auto=format&n=YnMJfeuar7Le0tPd&q=85&s=62b1dd081248147efb2c8e1ad3e543f2" alt="Selecting calls for a manual run" width="2048" height="1231" data-path="images/kpi-manager/manual-rca-select-calls.png" />

## Running one

<Steps>
  <Step title="Filter the calls">
    Choose the agent and filter the call list by date range and any other available criteria.
  </Step>

  <Step title="Select the calls">
    Review the results and select the calls you want analysed. The count is shown as you select.
  </Step>

  <Step title="Set the run's parameters">
    Select **Run RCA**, then set:

    * **Target (%)** — the value to measure this run against.
    * **RCA Model** — the model for this run.
    * **RCA Prompt (optional)** — focus for this run only; it does not overwrite the agent's saved prompt.
  </Step>

  <Step title="Read the report">
    The run starts in the background. When it finishes, the report appears in **History → RCA**, tagged **Manual**.
  </Step>
</Steps>

<Info>
  Manual runs work regardless of the **Auto RCA** setting, which makes this the right tool for investigating a specific complaint, testing a prompt idea before committing to it, or analysing a single campaign.
</Info>

<Tip>
  A manual run reads through every call you selected, so it takes a few minutes to finish. You can leave the tab and come back — the report will be waiting in History → RCA.
</Tip>