Skip to main content
Data Definitions (also called taxonomies) define the structure of data you want to extract from documents. They provide the schema that AI models and human reviewers use to classify and extract information.

​
What are Data Definitions?

A data definition is a hierarchical taxonomy that describes the fields, types, and relationships of the data you want to extract. For example, an invoice data definition might include:
  • Invoice Number (text)
  • Invoice Date (date)
  • Line Items (group)
    • Description (text)
    • Quantity (number)
    • Unit Price (currency)
    • Total (currency)

​
Managing Data Definitions

​
Viewing Data Definitions

The data definitions grid shows all taxonomies associated with your project. Click a taxonomy to drill into its taxon hierarchy.

​
Editing Taxons

Within a data definition, you can:
  • Add new taxons (fields) to the hierarchy
  • Configure taxon properties (data type, validation rules, extraction hints)
  • Rearrange the hierarchy by dragging taxons
  • Delete taxons that are no longer needed
For a comprehensive guide on building data definitions, see the Data Definitions guide and the Taxonomy Concepts documentation.