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

# Audit

> Drill into a single call's evaluation in detail — transcript, evidence snippets, and per-parameter scoring.

## Overview

While [Monitor](https://docs.neuroncx.in/assessment/monitor.md) and [Analyse](https://docs.neuroncx.in/assessment/analyse.md) show aggregate performance, **Audit** is where you review one call's evaluation in full detail — the actual evidence the AI used to arrive at a score.

***

## 1. Auditing Individual Calls

To review a single call's evaluation in detail, use **Cx Ops → Audit**.

**1. Go to Cx Ops → Audit**

Select the audio/call icon at the top to filter for audio-based interactions.

**2. Review the Audit list**

This view shows a summary (Agents, Topics, Languages, Forms), an Exclusion toggle, and a results table with per-schema score columns alongside an auto-generated call summary.

![Auditing Individual Calls](https://mintcdn.com/neuroncx/jlFg-zR2lq_hzngy/images/auditing_individual_calls.png?fit=max\&auto=format\&n=jlFg-zR2lq_hzngy\&q=85\&s=7af6fb0bdb7626af146e5e6ca9a9722e)

### 1.1 Call Detail Drill-Down

Clicking into a call opens its detail view, split across four tabs:

> [tab] Audit
>
> Overview of the interaction — overall score, schema-level score summary, the AI Config/prompt used, and the call log with playable audio recordings.
>
> ![Audit](https://mintcdn.com/neuroncx/jlFg-zR2lq_hzngy/images/audit.png?fit=max\&auto=format\&n=jlFg-zR2lq_hzngy\&q=85\&s=a450fe8db7c0c25e6419c5a433123787)

> [tab] Transcription
>
> A waveform audio player, an auto-generated call **Summary**, and the full speaker-labeled transcript.
>
> > [!NOTE]
> >
> > If the original call was in a language other than English, the transcript shown here is automatically translated into English.
>
> ![Transcription](https://mintcdn.com/neuroncx/jlFg-zR2lq_hzngy/images/transcription.png?fit=max\&auto=format\&n=jlFg-zR2lq_hzngy\&q=85\&s=839b5fc10e8e17695c7481481b628195)

> [tab] Snippet
>
> Evidence for each parameter — grouped by schema, showing the exact reasoning/snippet the AI used, its confidence percentage, and the resulting score (or true/false for flags).
>
> ![Snippet](https://mintcdn.com/neuroncx/jlFg-zR2lq_hzngy/images/snippet.png?fit=max\&auto=format\&n=jlFg-zR2lq_hzngy\&q=85\&s=a9a6adcdaff78370aeebd21c20428b1c)

> [tab] Scores
>
> A clean table — **Form | Parameter | Max Value | Score | Percentage** — covering every schema/parameter scored for that call. Scores are manually editable via the pencil icon, for override where needed.
>
> ![Scores](https://mintcdn.com/neuroncx/jlFg-zR2lq_hzngy/images/scores.png?fit=max\&auto=format\&n=jlFg-zR2lq_hzngy\&q=85\&s=6a9b668ab2d5622d2c07b370d5b6e30f)

***

## Next Step

Continue to [Analyse](https://docs.neuroncx.in/assessment/analyse.md) to break scores down by agent and skill dimension.

***


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.