Skip to main content

Documents and automatic extraction

The request's Documents tab combines the catalogue, uploaded files, classification, extraction status and access to the extracted values.

Documents with automatic extraction enabled and an AI-classified PDF
The document row shows its type, origin, processing result and review actions.

Upload processing​

With Extract automatically selected, uploading a PDF starts a durable processing flow:

  1. the content is used to recognise the document type;
  2. the workspace reuses a matching type or creates one;
  3. a matching extraction schema is reused or prepared;
  4. extraction starts and its state appears on the row and in Checks.
The upload area and a completed automatically processed document
AI detected identifies the classification source; View extraction opens the evidence.

Use Change type when the detected type is wrong. Use Read again after changing the type or schema. Clear Extract automatically before upload when you want to stop after classification.

Registry catalogue​

The catalogue is listed automatically and each item shows its price. Ordering remains explicit and is confirmed one document at a time. Delivered files are copied into workspace storage so Download keeps working after a register link expires.

The document order confirmation with the selected filing and price
The price is confirmed before the document is ordered.

A purchased PDF enters the same processing flow as an upload. Its type is supplied by the register when available, and extraction starts automatically while the option remains enabled.

Deleting and reprocessing​

Deleting an upload removes its file and extraction from the request. A failed row shows Retry processing. The Checks panel provides the same failure and progress state without requiring the Documents tab to remain open.

In this section​