Skip to main content

Document Processing

When a document enters Talonic, it passes through the processing pipeline: OCR converts scanned pages to machine-readable text, the 529-type ontology classifies the document type, and the Field Registry identifies which fields to extract. This preprocessing stage feeds directly into the four-phase extraction pipeline, ensuring every document is properly parsed before AI extraction begins.

Related

Supported Formats
File formats processed by the pipeline.
Document Types
The 529-type classification ontology.
Four-Phase Pipeline
Resolve, Agent, Validate, Re-read phases.
Phase 1: Resolve
Document classification and type detection.