Monitor reports

Documentation Index

Fetch the complete documentation index at: https://docs.superbryn.ai/docs/llms.txt

Use this file to discover all available pages before exploring further.

​

What are monitor reports?

Monitor reports aggregate audit results across all analyzed calls for an agent. Instead of reviewing calls one by one, you get a high-level view of quality patterns grouped by user intent and verdict.

​

Scoping with labels

Before generating a report, you can scope it to a subset of calls using labels.

• Leave the scope empty to include all audited calls
• Combine one or more labels to narrow the report
• Quick-scope cards show each label with its call count and cache status (fresh, stale, or missing)

​

Generating a report

1. Go to Monitor Report for your agent
2. Optionally select labels to scope the report
3. Click Generate Report

Reports are cached after generation. If no new calls have been ingested since the last run, the cached report is served instantly without additional LLM cost. If new calls exist, you can regenerate to include them.

​

Cache status

​

Browse by verdict

The default view shows a 2x2 grid of verdict cards: Each card shows:

• Number of calls with that verdict
• Number of unique user intents
• Issues identified
• Percentage of total calls

Click a verdict card to see all user intents under that classification.

​

Browse by intent

The intent view shows a pivot table with user intents as rows and verdicts as columns. This helps you identify which intents have the highest failure rates. Click any cell to drill into the specific calls for that intent-verdict combination.

​

Drill-down views

When you click into a verdict or intent, you see:

• Grouped issues - common problems across calls in that slice, with call counts
• Call IDs - click any call to open the full call analysis panel

This lets you trace from a high-level quality pattern down to the individual call that caused it.