LoanLogics released a whitepaper titled, “Big AI Driven by Today’s Machine Learning,” which provides a deeper understanding of how artificial intelligence (AI), and machine learning specifically, can help mortgage lenders reduce and even eliminate redundant, repetitive tasks while creating data purity and improving the borrower experience, the company announced in a release.

According to the whitepaper, the best uses of AI and machine learning technologies within the mortgage industry reside in data and document processing. The authors explain how Capture 2.0 technologies, also known as “automatic identification and data capture” and “intelligent capture,” are able to leverage machine learning tools to determine the accuracy, quality and completeness of loan data as it is being collected from the borrower, the company said.

Ultimately, Capture 2.0 technologies can reduce the costs not only of manually retyping data, but also eliminating the need to ask borrowers to resubmit information.

However, according to the release, Capture 2.0 technologies require a robust set of document examples that make enough “training data” available to maximize the benefits of machine learning. When sufficient data is available, machine learning tools can classify loan documents and extract and validate loan data with much greater speed and accuracy than traditional optical character recognition (OCR) technologies used in the industry today.

“While OCR is effective, the many different types of documents and the locations of data on those documents can create challenges for data extraction,” LoanLogics Executive Vice President and Chief Business Officer Craig Riddell said in the release. “By combining OCR with automated document recognition (ADR) and automated data extraction (ADE) tools, Capture 2.0 technology takes data capture and classification to a new level.

“Not only can lenders process mortgages at lower per-loan costs, but they alleviate the No. 1 borrower frustration of constantly resubmitting forms and documents throughout the process. It’s a win-win situation,” Riddell added.

The whitepaper explains how machine learning is a starting point for lenders to be able to realize immediate savings, while maximizing the potential of more advanced AI applications. It also explores where AI and machine learning will take the mortgage industry, and what lenders must do to take advantage of these innovations for themselves, the release stated.