Before we can extract data or do much of anything with our documents, we first need to have text data on our documents. No matter if the document has native text or is image based, we can obtain text through the Recognize Step. Traditional OCR was great at getting text data off of documents, but there were certain areas where it struggled. Not only is AI OCR easier to set up and more convenient, but it also can vastly improve OCR results.