Available column types
Each column in a Collection has a type that determines what kind of data it holds and how the AI generates its content.
Column type | Description | Example |
Text | Free-form text of any length | Contract summary, risk analysis, key clauses |
Number | Numeric values with formatting options | Rent amount, total revenue, page count |
Boolean | True/False values | Has renewal clause, requires approval |
Selection | One or multiple values from a predefined list of options (single-select or multi-select) | Risk level (Low / Medium / High), Tags (Urgent, Review, Approved) |
File | An uploaded document, image, audio, or video | The source lease agreement, a product image |
JSON | Structured data in JSON format | Contact details, address, metadata |
Date | Date or date/time values | Lease start date, deadline, signature date |
Web | URL / website link | Company website, reference link |
Email address | Contact email, signatory email |
Number format options
Number columns support several display formats:
Format | Example |
Plain | 1234.56 |
Number with Commas | 1,234.56 |
Currency | $1,234.56 |
Percentage | 12.34% |
Date format options
Date columns support several display formats:
Format | Example |
Date | 2026-02-08 |
Date & Time | 2026-02-08 14:30:05 |
Date & Time (ms) | 2026-02-08 14:30:05.123 |
Time | 14:30 |
JSON column
The JSON column type lets you store structured data with a defined schema. When creating a JSON column, you can provide a JSON Schema that specifies the structure the AI should output.
Built-in schema templates:
Contact — Name, email, phone
Address — Street, city, state, zip, country
Product — Name, price, currency, stock status
Metadata — Tags, category, priority
You can also define a custom schema for any structured data format.
AI Autofill
When creating a column, you can enable AI Autofill. With AI Autofill turned on:
You provide a column name and a description of what to extract.
The Agent automatically generates the detailed extraction instruction.
When documents are uploaded, the Agent populates the column for every row.
With AI Autofill turned off, the column must be filled manually.
The first column
The first column in every Collection has a special role. It defines the primary entity — typically the uploaded file (e.g., a contract, report, or invoice). All other columns are generated based on this first column's content.
Adding and editing columns
Click Add a column in the Tabular View to create a new column.
Click a column header to edit its name, type, description, or AI instruction.
Drag columns to reorder them.
Right-click a column header for options like delete, hide, or duplicate.