Scanned documents, where data is embedded in images, must be turned into digital form before they can be processed by the software robot. For this SmartRPA uses the Kofax Transformation Module, which is an industry leading OCR extraction engine built upon the best OCR engines in the market, combined with intelligent AI technology for accurate extraction.
Kofax KTM can use traditional rule-based extraction logic which works well when there are few variations over many scanned documents.
For more complex documents, an intelligent AI engine can be taught by feeding it a number of samples to learn from. The more and the better samples, the more accurately it can extract data – even in new situations. Outliers will automatically be forwarded for human check and any feedback from this process is used to further train the OCR engine.
SmartRPA Kapow robots can use the Kofax KTM product to automatically overview an input queue of scanned documents and process the data when extracted successfully.