As shown in the graphic on the left, BR eliminates or substantially compresses steps often taken in traditional document processing. BR’s visual classification technology automatically groups visually similar documents – even those with no text. From the first day that processing begins, BR provides project or process managers with a fully-informed awareness of what document types are present and in what quantities.
Document coding is as simple as clicking and dragging with a mouse on a document image to indicate data elements of interest – the extracted data elements are instantly displayed in a grid showing how the extraction operates. BR provides expressions to qualify the data to be extracted and provides filters to specify how the data elements will be formatted.
Visual classification and visual coding are both consistent and persistent, making QC efforts far more manageable with fewer people. BR notifies operators when new document type clusters form so that they can set processing rules for them. With BR’s unique global glyph catalog, BR can recode previously processed documents to extract further data elements without having to completely reprocess them – that’s called repurposing without reprocessing.
BR’s glyph recognition system provides single-instance error correction – editing a single instance of an error corrects all instances of that error – no more tedious, linear, instance-by-instance editing of errors. Perhaps best of all, BR’s processing can be done in parallel, as shown in the following graph. From the very start, teams can be working in parallel to label clusters, define extraction rules and edit text.
In sum, BR provides quicker starts, quicker finishes, and better data than traditional document processing approaches.