> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs.ada.cx/mcp/prompt-library/deep-dive-analysis/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.ada.cx/_mcp/server.

# Deep-dive analysis

Combine multiple insights for comprehensive analysis and recommendations.

| Prompt                                                                                                                                                        | What you'll learn                                                          |
| ------------------------------------------------------------------------------------------------------------------------------------------------------------- | -------------------------------------------------------------------------- |
| "Review 100 inquiry summaries from yesterday where CSAT was low and share recommendations on what knowledge, coaching, or playbooks we should add or update." | Actionable recommendations based on recent low-satisfaction conversations. |
| "For all conversations today that weren't resolved by our AI Agent, what were the most common reasons for failure?"                                           | Pattern analysis across unresolved conversations.                          |
| "Compare our AR this month vs. last month and analyze what might be driving any changes."                                                                     | Month-over-month analysis with potential root causes.                      |

> **Tip**
>
> **Tips for better results:**
>
> * Be specific about timeframes (for example, "last 7 days", "yesterday", "this month").
> * Include the number of conversations to analyze (for example, "Review 50 summaries...").
> * Ask follow-up questions to dig deeper into initial findings.
> * Use available filters: CSAT scores (1–5), resolution status, topics, playbooks, language, channel, browser, device, and more. See [`get_available_filters`](/mcp/tools/get-available-filters) for the full list.
> * When creating visualizations, ask follow-up questions to refine the chart (for example, "adjust the y-axis to start at 30%").