We have been testing this for several months and are so proud to be releasing such a powerful feature to our clients. We have always automatically corrected common OCR errors, now this has been significantly improved by utilising all data points on a page to infer, identify and amend OCR errors you may experience, after the raw data extraction has taken place.
For example, if a hole-punch obscures two date values, on every page, this can be time consuming to locate and correct manually. StatementReader now reviews and identifies errors; then (using background shading in the Excel output so the user remains in control), uses the surrounding dates to correct the errors, or if this cannot be done ‘safely’, the potential error is flagged to the user.
The validation of dates and amounts is crucial to obtaining a valuable analysis of the extracted data. The completes the three step process of an optimised data process - from extraction, automated validation to generating a one-click analysis report in Excel.