The Scenario
The product team shipped a new onboarding assistant two weeks ago. Your job this week is to review the first 80 text chat conversations it had — not the voice calls, the text chats — to see how it is actually handling new users before you decide whether to expand the rollout.
The conversations are in Vapi. Your analysis tool is an Excel workbook.
You find the Chats section in the Vapi dashboard and see a list of conversations. To get the data you need — chat ID, session ID, creation date, total cost, message count — you would need to click into each conversation individually. There is no obvious export.
The bad version:
- Click into the first chat, note the ID, session ID, date, cost, and message count
- Go back to the chat list, click the next one, repeat 79 more times
- Paste everything into your workbook row by row, lose your place twice when the page reloads
- Realize you forgot to note the session ID for 12 of the chats and have to go back
You are supposed to have a preliminary read on these conversations before Friday's retro. Data collection should not be the hard part.
The Easy Way: One Prompt in SheetXAI
SheetXAI is an AI agent that lives inside your Excel workbook. It connects to Vapi through its built-in integration and pulls the full chat log for a specific assistant into your worksheet in one operation.
Export every Vapi chat from the past 14 days into my Excel sheet with columns for chat ID, assistant ID, session ID, creation date, and total cost
What You Get
- One row per chat with chat ID, assistant ID, session ID, creation date, and total cost in separate columns
- The full list populates without any dashboard navigation or manual copying
- The workbook is ready to sort by date, filter by cost, or flag by message count for further qualitative review
What If the Data Is Not Quite Ready
You only want chats for a specific assistant
Export every Vapi chat for assistant ID 'asst-abc123' from the past 14 days into my Excel sheet with columns for chat ID, session ID, creation date, cost, and message count
You want to flag unusually short conversations that might indicate failed sessions
Export every Vapi chat for assistant ID 'asst-abc123' from the past 14 days into my Excel sheet with columns for chat ID, creation date, cost, and message count — add a flag column marking any chat with fewer than 3 messages as 'possible failure'
You need chats from multiple assistants in the same workbook
Export all Vapi chats from the past 14 days for assistant IDs 'asst-abc123' and 'asst-def456' into my Excel sheet with columns for chat ID, assistant ID, session ID, creation date, cost, and message count — one row per chat
Pull, score, and flag all in one shot
Export all Vapi chats for assistant ID 'asst-abc123' from the past 14 days into my Excel sheet with columns for chat ID, session ID, creation date, cost, and message count — flag chats with more than 20 messages as 'long session', fewer than 3 messages as 'short session', and leave the flag blank for chats in between
Combining the pull and the segmentation logic in one prompt means the workbook arrives ready for qualitative review without any extra column work.
Try It
Get the 7-day free trial of SheetXAI and open an Excel workbook before your next assistant review cycle, then ask it to export the Vapi chat log for the assistant you are evaluating. For related tasks, see how to export the full call log or run an analytics query.