Custom information extractor: define your own fields

Your documents, your schema

Every business has documents nobody else has: proprietary order forms, industry certificates, partner-specific paperwork. The custom information extractor lets you define an extractor that pulls exactly the information you choose from any custom document, text, or message.

How you define it

1. Name your fields: describe the values you need (identifiers, dates, parties, amounts, free-text clauses).
2. Point it at documents: PDFs, images, plain text, or message bodies, in any major language.
3. Get structured output: your schema back as clean data, with a confidence score per field.

Typical uses

• Proprietary order and requisition forms
• Industry-specific certificates and permits
• Contract clauses and terms capture
• Anything the prebuilt catalog does not cover

Works alongside the catalog

Use prebuilt extractors for standard paperwork (invoices, POs, bills of lading) and custom extractors for the documents unique to your operation. Both return the same structured shape, so downstream automation treats them identically.