IDP (Intelligent Document Processing) is the technology that combines OCR (optical character recognition), NLP and machine learning to extract structured data from free-format documents — invoices, receipts, contracts, insurance policies, forms. It is one of the areas with the clearest ROI in current corporate implementations.
The difference between IDP and pure OCR is significant. OCR converts an image into text; IDP understands the text. On an invoice, OCR says 'there is a number 1234.56 in this position'; IDP says 'this is the total amount of this invoice, corresponds to supplier ABC, and should be posted to accounting account X'. The difference is between syntactic accuracy and semantic understanding.
Document Management by INFOS (via Tungsten Automation) implements IDP with accuracy above 95% on typical Portuguese documents — invoices, receipts, transport notes. Approval flows are triggered automatically with the extracted data; cross-validation with the ERP (does the tax number exist? is the amount consistent with the purchase order?) catches errors before they reach finance. Invoice processing time typically falls from days to hours.