Editor's note: This story originally appeared in the November edition of DSNews, out now.
The COVID-19 pandemic will likely be a cross-roads and accelerator to the long-term future for many industries. In the last few years, we started to check into hotel rooms with our phone without having to stop at the front desk. Restaurants provided modules to help customers pay without handing their credit cards to the waitress. Minimizing personal contact has likely become a trend which will accelerate innovation more than we can imagine.
In our own industry, we have had electronic access to necessary loan information for decades. Innovations in our industry have come from many different areas: property preservation companies, technology firms, law firms, title companies, collection agencies, claim filers, and auction companies, as well as banks and non-bank servicers. Innovations came as we integrated information among disparate systems, providing data and documents to facilitate the process.
How the Industry Innovates
Invoice processing technology is a great example which changed our world in the last quarter century. Logical milestones resulted in automation to a process which, 25 years ago, was littered with multiple requests for payment, stacks and stacks of paper, duplicate payments, supplemental claims, inevitable claim-filing errors, and extrapolations of audit findings. Much of that has been cured by the invoice processing solutions of 20 years ago.
Similarly, property preservation and inspection companies put technology in the field years ago that eliminated phones having to be connected to paper reports. Problematic issues at a property and necessary work for conveyance or to cure a violation are bid virtually in real time and returned to the servicer immediately.
Technology has provided us with an amazing ability to communicate to borrowers through web portals, emails, text, and form letters, and allowed borrowers to be able to navigate through voice attendants and receive answers via phone or online in order to understand what is happening with their accounts. This includes up-to-the-minute information regarding how payments were applied, tax, and insurance information. Incredible strides have been made in the transfer of data and documents between loan servicers and their related service providers. Attorneys serving the mortgage industry have similarly established these protocols with their clients, title companies, process servers, auction companies, and, most importantly, with the courts. The process is now paperless, allowing for enormous amounts of information to be processed and transmitted.
The pandemic further accelerated this process, albeit with some twists. Within weeks, we all learned how to work at home and overcome whatever fears and obstacles remained regarding information security. We determined how to put borrowers into forbearance plans without them having to talk to an agent or provide financial information. This was aided by the Cares Act, which facilitated the little or no documentation approach to these issues and allowed servicers to provide service to borrowers impacted by the pandemic.
In a world where cars are driving themselves and algorithms drive everything from stock trading to where baseball teams position their infielders (I’m not a big fan of the shift), a logical answer is the continued innovations afforded to us by machine learning (ML) and artificial intelligence (AI).
Servicers are striving to determine what factors cause servicing errors. Attorneys and other vendors are as well. Everyone is trying to understand the dynamics of scaling up after scaling down in an uncertain world. Some of the factors contributing to errors we can list easily (service transfers, copied and scanned and imperfect documents, loan history that indicates a possible payment anomaly). The industry has access to hundreds of data fields and documents that may hold answers to questions and possible process improvements. It’s just a matter of time before we determine that some aspect of the servicing transfer or keywords in borrower communication identify opportunities for the same kind of lift and improvement that the industry received from invoice automation. That timeline may well be accelerated through the utilization of AI to assist in identifying anomalies that have resulted in previous losses.
Bob Caruso, CEO of ServiceMac, has been a leading innovator in the mortgage servicing industry for more than 30 years. Caruso shared that his team has built a proprietary quality assurance software that incorporates the database, rules, and workflow which will ultimately be the foundation for ServiceMac to effectively leverage AI in the future.
“Artificial Intelligence has reshaped how we uncover data issues that contribute to compliance risk and borrower dissatisfaction,” said Gagan Sharma, President and CEO of BSI Financial Services and Founder of Bizzy Labs, a regtech company that employs artificial intelligence in its Libretto engine. “Using AI, we can scour loan files for data exceptions more thoroughly and frequently, helping servicers identify issues on their portfolio. This big data approach to perfecting loan data files and pools will disrupt traditional servicing business models by elevating the user experience for lenders, investors, and borrowers.”
One of the biggest areas of concern for attorneys is the Personally Identifiable Information (PII) that is regularly contained in the documents which are attached to complaints, proofs of claims, judgments, and other documents. Redacting this information from documents is a time-consuming, tedious, monotonous, and imperfect process. Some of the technology provided to assist with the process is imperfect, allowing boxes to be moved unless a document has been appropriately flattened, or even reprinted and copied. Firms across the country have been engaged in projects to assist with painful after-the-fact fixes to cure the situation, including court hearings, notification of the consumer, and possible sanctions and financial penalties for the servicer and attorney. With OCR and ML technology, however, the opportunity exists for document redaction, which is manual and imperfect, to be much more consistently correct and significantly less costly.
After the Moratoria Lapse ...
Continue reading on p. 61 of the November issue of DS News, available here.