Intelligent document processing (IDP)
Definition
Intelligent document processing extracts structured data from invoices, delivery notes, contracts or forms, validates it against master data and posts it into the target system. It combines OCR for reading with models for interpretation and rules for verification.
Reading the page is the easy part. The value lies in the checks: does the supplier exist, does the amount match the order, is the tax treatment correct, is this a duplicate?
A well-built pipeline reports its own confidence and routes only uncertain documents to a human.
In practice
- OCR plus model-based interpretation plus rule validation
- Confidence scores decide automatic posting or review
- Duplicate and master-data checks prevent expensive errors
Related terms
- OCR (optical character recognition)OCR converts scans, photos and PDFs into machine-readable text. It is the entry step of document automation and delivers characters only — the meaning, validation and posting logic are added by the layers above it.
- Human in the loopHuman in the loop means a defined point at which a person reviews, approves or corrects the AI system's work. The handover is designed in advance for specific triggers — value thresholds, low confidence, unknown counterparties — rather than left to chance.
- ERP integrationERP integration connects operational software and AI agents to the system of record for orders, articles, customers, stock and accounting, so automation reads current data and writes results back into the books instead of into a parallel spreadsheet.