Collected Data: reviewing what your agents captured
Every field your agents capture lands here as a row. How flagged fields are decided, how to correct and verify one, and what the three export formats give you.
8 min readUpdated 26 August 2026
Collected Data in the sidebar is where the structured fields your agents pull out of calls end up. The page describes itself as structured data captured from your calls, for reviewing flagged fields and exporting.
It is the other half of the Data capture panel on an agent. That panel decides what to collect. This page is where you read it, fix it and take it away.
Nothing here yet?
An empty page says so plainly and tells you the cause: configure capture fields on an agent, on its Data Capture tab, and rows appear here after each call.
So an empty table almost never means the page is broken. It means either no agent has capture fields set up, or no call has happened since they were.
What a row is
One row is one call's captured data. Four fixed columns describe the call, and every field your agents capture becomes a column of its own alongside them.
Tip
Your own fields appear as columns automatically, named by the key you gave them in the Data capture panel. That is why short, readable keys are worth the thought at the time.
| Column | What it holds |
|---|---|
| Date | When the call happened |
| Agent | Which agent captured it |
| Source | How the call arrived |
| Status | Needs review, or Verified |
Flagged fields
The subtitle promises flagged fields, and the rule behind the flag is worth knowing rather than guessing at.
A field is flagged when the agent was not confident about it, which in practice means a confidence below 60 percent. Flagged fields are outlined in amber so they stand out inside an otherwise ordinary-looking row.
This is the feature that makes the page usable at volume. You are not proofreading every call. You are looking at the handful of values the agent itself was unsure about, which is where the wrong postcodes and mangled surnames actually live.
Correcting and verifying a row
Open a record and its fields become editable. Two ways to finish, and they mean different things.
- Save keeps your correction and leaves the row as it was, still Needs review.
- Saving with verify marks the row Verified, which is how you record that a human has actually looked at it.
Tip
Filter to Needs review, work down the list, verify as you go, and the filter empties itself. That is the entire review workflow and it is worth doing daily rather than weekly.
Filtering
Three filters sit next to the record count, and the record count updates as you narrow.
Tip
Source is the one people forget. Filtering to Campaign separates the calls you initiated from the ones that came to you, which are rarely worth reading together.
| Filter | Options |
|---|---|
| Agent | Any of your agents |
| Status | Needs review, Verified |
| Source | Web call, Inbound, Single outbound, Campaign, Unknown |
Choosing your columns
The Columns control on the right of the filter bar carries a count and opens a picker. Every column can be turned off, including the four fixed ones, and the columns are resizable by dragging the right edge of a header.
Both the choice and the widths are remembered in your browser, so the layout you set survives navigating away and coming back. That also means it is per browser rather than per account, and a colleague on their own machine sees the defaults.
Exporting
Export at the top right opens a short menu with three formats.
| Format | What the menu calls it |
|---|---|
| CSV | Spreadsheet |
| XLSX | Excel workbook |
| Document |
The export respects whatever filters are active, which makes it more useful than it first looks. Filter to one agent, one source and Verified only, and the file you get is exactly that slice rather than everything you have ever collected.
Careful
Exporting before reviewing hands the unverified values downstream along with the good ones. If the file is going into a CRM or to a client, filter to Verified first.
Common questions
- Collected Data is empty. Why?
- Either no agent has capture fields configured, or no call has happened since they were added. The empty state points at the Data Capture tab on an agent, which is where fields are defined.
- What makes a field flagged?
- Low confidence. A value the agent was less than 60 percent sure of gets flagged and outlined in amber so you can find it without reading the whole row.
- What is the difference between Needs review and Verified?
- Needs review is the state everything arrives in. Verified is what you set once a person has checked the row, and it is the filter that lets you separate checked data from raw data later.
- Can I fix a wrong value?
- Yes. Open the record, edit the field, and save. Saving with verify does both at once.
- Where do the column names come from?
- From the keys you set in an agent's Data capture panel. Change a key there and a new column starts, with older rows keeping the old one.
- Do my column choices apply for everyone?
- No. Visible columns and column widths are stored in your own browser, so each person sets their own view.
- What formats can I export?
- CSV, XLSX or PDF. Whichever you pick, the file contains what your filters are currently showing.
Was this helpful?