This video demonstrates a very complex process of indexing and classifying a high volume of multi-page mortgage documents, which are in an unstructured format, into 200+ categories/document types. The process uses Robotic Process Automation (RPA), Optical Character Recognition (OCR), Artificial Intelligence (AI), Machine Learning (ML), and Natural Language Processing (NLP).
It also elucidates how the ML logic gets strengthened by handling the exceptions and training the algorithms to execute the process better in an unsupervised learning scenario over a period of 6 months.
The video essentially demonstrates the functioning of an unattended bot using AI engine to index and classify unstructured documents into 200+ different categories/document types in a secure environment. It also shows how the documents are then exported to a Document Management System built using FileNet while maintaining an audit log of all the sub-processes involved.