For modern-day loan package processing, manual document management just doesn’t cut it.
Mutually Reinforcing Challenges
With loan document packages often exceeding 500 pages and potentially including over 100 unique document types, the mortgage loan origination lifecycle is heavily dependent on accurate and efficient document classification. Loan documents come from multiple sources, including brokers, lenders, borrowers, employers, and online vendors, and the set of documents required varies from state to state and county to county. Inconsistent and potentially confusing nomenclature can make it difficult to categorize documents by hand. For example, “1003”, “mortgage application form”, “uniform residential loan application”, and “URLA” all refer to the same document.
Furthermore, the PDF file created by combining scanned images of these documents doesn’t adhere to a single, standardized document order. Over the course of the loan origination lifecycle, document management teams will typically encounter multiple versions of many of these documents, too. As you can see, a competitive title company or document preparation firm cannot rely on manual efforts to manage documents in an accurate, timely, and cost-effective manner. For these reasons, loan servicers, title companies, and other organizations in the mortgage industry must instead turn to automated IT solutions.
OCR to the Rescue
One fundamental tool used for automatic document indexing is optical character recognition (OCR), which converts the pixels that make up an image of a page of text into a stream of characters that computers can interpret and store. OCR allows indexing software to automatically categorize software, group multiple versions of the same document, and extract information to be used by other software systems.
While some OCR vendors like Kofax, Xerox, and Paradatec offer generic character recognition capabilities that lack any specific knowledge of loan documents, whereas Visionet Systems uses OCR engines that have been trained using the various types of loan documents, encountered while indexing a loan package. This cloud based accelerator platform hosts solution suite with a powerful combination of NLP (Natural Language Processing), analytics, domain specific knowledge and OCR technology elements to delight customers with superior accuracy rates as compared to their generic counterparts. Armed with this domain-specific knowledge of document layouts and the types of information to expect in each form field, expert systems for loan document indexing significantly improve Cost, Quality and Time parameters of document indexing process.
VisiLoanReview: Visionet’s Auto-Indexing Solution
Given the various challenges to loan document management mentioned above, Visionet has put a lot of thought into designing VisiLoanReview (VLR), a solution for indexing and validating loan documents. The embedded OCR engine separates loan package PDF files into individual documents, which are then automatically classified and organized. VLR maintains a complete revision history for each document, making it easy to view and use information from the most current version. Once automatic indexing is complete, VLR highlights the exceptions, which can be manually checked to ensure the exceptions are taken care of appropriately. This entire process significantly cuts down the processing cost of a mortgage loan. It is a win-win solution for all the stakeholders: borrowers can get faster loans, brokers and lenders can get higher accuracy in lesser time and can manage different types of documentations predictably.
The encouraging 90%+ accuracy rate achieved by VLR with loan documents has led Visionet to repurpose this auto-indexing engine so that it can be used with other document types across all kinds of business processes. In the near future, Visionet’s auto-indexing technology will be offered for managing invoices, requisition forms, insurance documents, contracts, and many other complex business documents. These advanced technologies will help organizations streamline operational processes, cut costs, improve accuracy, and save time, resulting in a substantial competitive advantage.