Flexicapture Demo - Abbyy

To appreciate the "wow factor" often present in a FlexiCapture demo, one must first understand the limitations of its predecessors. Traditional OCR technology was rule-based and brittle. It could convert an image of text into machine-encoded text, provided the document was a fixed form—like a standardized test sheet. However, if a business received an invoice from a new vendor with a unique layout, traditional OCR would fail. It did not know where to look for the "Total Amount" or the "Invoice Date" because those coordinates changed from page to page.

Unlike legacy systems that required hard-coded templates for every single vendor or document type, FlexiCapture uses a "trainable" approach. The demonstrator might show an invoice from "Vendor A" and "Vendor B." Vendor A puts the total at the bottom right; Vendor B puts it at the top left. FlexiCapture does not memorize coordinates; it understands context. It looks for keywords like "Total," "Amount Due," or "Grand Total" and interprets the data associated with them based on relative positioning. abbyy flexicapture demo