Mortgage company eases loan processing with OCR
Vintage Mortgage Group turned to technology to automate processing the many forms it handles each day, helping homebuyers purchase property in the San Francisco Bay Area.
The company, which provides consultation and loan origination services, has helped more than 14,000 families purchase homes. However, the system required manual processing of dozens of forms each day from up to 15 different lenders and in that many formats. Vintage Mortgage Group received the loan applications and related documents as PDF files, but even digital loan documents required manual processing.
Alex Gonzalez, CEO at Vintage Mortgage Group, explains, “Each of the 15 banks we use has its own format and wording for the documentation required by Fanny Mae and Freddie Mac. Plus, we have to keep up with ever-changing government regulations. For security, the PDFs from each bank were locked so we had to hand-type every document into Microsoft Word to create a loan origination form that complied with each bank’s internal rules and the latest regulations.”
Vintage Mortgage Group wanted to automate the inefficient, time-consuming system for loan processors while maintaining the volume of home sales closed each month.
Vintage found a partial solution with loan origination software (LOS) that auto-populates loan forms created in Microsoft Word, each customized to an individual bank. But the solution couldn’t read PDF files, resulting in loan processors needing to manually create documents for every one of the 15 banks.
To address the situation, Vintage Mortgage Group chose an optical character recognition (OCR) solution, ABBYY FineReader, which changes PDFs into Word docs accurately. Now loan processors simply take a PDF, use ABBYY FineReader to turn it into a Word file and then use the LOS solution to automatically populate a form with the client’s details.
According to ABBYY, the software is so accurate that Vintage Mortgage Group can create customized forms with little need for correction. Automating the ingestion of accurate data from PDFs into the LOS has saved Vintage Mortgage Group hours per loan file.
Gonzalez says, “FineReader saved me hours right off the bat. What normally would have taken me half an hour took me three minutes.”