Documents

The Documents tab is the default view on a dataset's detail screen. It is where you browse every published record in a dataset, one card or row per document, and choose how those records are displayed.

Clicking a dataset row on the Datasets page opens its detail screen, and Documents opens first. It is one of up to three tabs for the same dataset — Documents, Responses, and Overview — each of which looks at the same underlying records in a different way. Documents is for browsing whole records; the other two are for analysis.

Why use the Documents tab?

Documents is your starting point for working with a dataset. From here you can:

• Browse every published record in the dataset in one place.

• Switch layouts between table, card, and list to suit what you are doing.

• Filter, search, and sort to narrow down to the records you need.

• Export the current view to CSV or JSON for analysis outside the platform.

Understanding the interface

The Documents tab has a consistent layout, top to bottom:

• A cover/header at the top, with a primary Import Data action for adding records to the dataset.

• An actions bar holding the Filter toggle, keyword search, view switcher, sort, and Download Data.

• A filter sidebar on the left (collapsible).

• The listing body, which shows your records in the chosen layout, a live total count ("Showing N results"), and loads more as you scroll.

Switching views

You can switch how records are displayed using the view switcher at the top right. By default you land in Table view.

There are three layouts to choose from:

Table records as rows in a table. This is the default.

Cards

records as a grid of cards, best for records with cover images.

List

records as a vertical list of wide rows.

Whichever layout you choose, each record shows the same core details: its title, cover image (where one exists), the contributor's name, and the publish date.

Reading a record

Click any record to open its individual document page, where you can read it in full.

Beyond the core details, records can carry a couple of status markers:

• An Annotated indicator (a green highlighter icon on the title) shows that a record has AI annotations.

• An Under Review / Under Edit tag shows that a record is currently going through moderation.

Filtering and searching

Filters are generated automatically from your dataset's template, so the options available reflect your own fields — categories, tags, author, publish date, and so on. You can select multiple values to build the exact slice you want.

A few things to know about filters:

• Some datasets have preset filters that are always applied and cannot be cleared.

• Your filter and search selections persist as you move between the Documents, Responses, and Overview tabs, and survive a page reload within the same session.

At the top of the tab, a keyword search lets you look across records, ordering matches by relevance. If a search returns nothing, you can clear it in one click to return to the full list.

Sorting

The sort control orders records by Latest or Oldest publish date. Some workspaces additionally offer Title A–Z / Z–A.

Bulk actions

In Table view you can select multiple records and act on them together:

• Run AI Annotation — generate AI annotations across the selected records.

• Delete — remove the selected records.

Downloading data

The Download Data button exports your records for use outside the platform. You can choose CSV or JSON, and the export respects your current filters, search term, and sort order, so you only get the slice you are looking at.

Tips for working in Documents

• Start in Table view to scan quickly, and switch to Cards when you want to see cover images and browse more visually.

• Filter first, then export. Because the export follows your current view, narrowing down before you download keeps the file focused.

• Use the status markers to spot which records already have AI annotations and which are still in moderation before you rely on them.