The evolution of intelligent technologies is ushering in a new era of enhanced operations for mortgage servicers — one led by intuitive workflow guidance, embedded expertise and a harmonious balance between technology and human decision-making. The convergence of natural language processing, machine learning and generative AI signals a future where technology amplifies human performance, which results in the improvement of homeowner support. In this article, we explore three ways the application of intelligent technologies can simplify mortgage servicing for both internal ops teams and homeowners.
1. Intuitive workflow guidance
Servicing is complex and requires great specialization of the technology, as well as the team members who use it. These back-office teams are often required to adhere to specific procedures dictated by the servicing scenarios they encounter. To make navigating both new and repetitive servicing scenarios simpler, servicing interfaces can guide users through the actions they need to perform.
Machine learning (ML) and natural language processing can enhance business decisioning processes and help reduce the steep learning curve employees must overcome before achieving competence in their role. Prompts provided by the system will help eliminate the need for employees to rely on sticky notes to navigate stumbling blocks they frequently encounter. Without having to make calls or memorize where to look for regulatory references, servicing teams can be alerted on the next steps to take once tasks are completed. Combined with role-based functionality, managers could reassign employees more easily, creating greater fungibility.
To date, the industry has experimented with robotic process automation (RPA), scripting, code-intensive workflow, but all these tools were not developed with mortgage servicing in mind. Technology should be able to support servicing teams with intuitive workflows that allow for configuration and easy-to-navigate interfaces that harness all the data and processes they need to complete their jobs effectively.
2. Embedded expertise
There are lots of nuances to loan servicing that take years to learn. Programming these into servicing workflows makes it simpler for employees to take the proper steps consistently, minimizing business risk.
For example, servicing systems should flag a loan in advanced default, where the borrower has missed several payments, that has reached or exceeded its maturity date without the employee having to search for that information in separate places to compare data points. Instead, the technology should be proactive, alerting servicing teams to the loan condition and recommending next steps. Servicing technology should also be a resource for teams to look things up when they need additional information. Advances in natural language processing make it possible to pose simple queries and quickly receive detailed loan information, as well as prompts on next steps.
This type of functionality lets employees get up to speed quickly, view suggested next actions or ask for help from the platform and focus their time where it’s going to make the most difference without having to wade through lots of data.
Modern servicing technology should give servicing teams the data they need, relevant to the problem they're trying to solve, and the guidance to help them complete tasks correctly. It should also give the servicer the configurability needed to effectively support specific policies and processes.
3. Giving humans the final word
Servicers more than ever are having to balance the needs of the end consumer with maintaining their service-level agreements (SLAs) with investors. They contend with issues that require time and experience to resolve, including deal structure complexities, heightened asset surveillance (especially as interest rates stay stubbornly high), a steady stream of consent requests and seemingly constant loan-level reporting to investors and regulators.
We think technology should play a stronger role in flagging items that will require an experienced employee to step in and provide the “human touch,” making those decisions that can’t be solved by technology alone. Areas such as loss mitigation, where the borrower and the servicer work together to avoid foreclosure, will continue to require specialists. These experts have valued knowledge of the financial industry and are best positioned to assist borrowers with the available options, using technology to support how those options were made.
No matter how automated our offices become, the role of technology is to put people first. In other words, technology should not only simplify processes, but it should give experienced employees more time to focus on exceptions that require expertise and guidance.
Perhaps the greatest value of ML-driven automation is that it can bring consistency to servicing. Each homeowner’s case could be attended to methodically, subjected to the same oversights, processes and instructions, thus reducing human error and subconscious bias.
This is the direction technology needs to go to better support servicers, their teams and the homeowners we all are ultimately working for. We have always emphasized that advanced technology should streamline workflows by empowering humans, and amplifying their performance so people can focus on what matters most.
It’s a future that’s possible now as natural language processing, machine learning and generative AI are being integrated into more enterprise tools to enhance onboarding, servicing transfers, loss mitigation and other servicing functions.
It’s important to make servicing simple, and the time to do that is now.