From our first implementation of PROSAR-AIDA in 1997, we have been the pioneer and leader in advanced OCR Technology. Our hybrid approach of machine learning and manual rules support blends artificial intelligence and robotics with manual review, where needed, to produce high quality data output to drive downstream decisioning systems.
On its own, OCR generates text from document images, creating nothing more than a collection of words and numbers that have no broader meaning. PROSAR-AIDA’s AI capabilities assemble words into phrases and phrases into sentences, bringing context to this ‘collection of text’. By understanding context, the solution understands and identifies distinct data elements, even of the same data type. As an example, in the mortgage market, the first page of a Note contains three distinct date values. PROSAR-AIDA reads the page just as a human would, then based on the phrasing used around each date instance, correctly identifies the date of the Note, the first payment date, and the mortgage maturity date.