With national mortgage rates at historic lows and homeownership demands rising among millennials , there is speculation that we could be looking at another "Roaring 20’s scenario"  when the coronavirus pandemic begins to recede.
That’s why there’s never been a more pressing need for mortgage originators to produce higher quality loans faster than ever before. Competition among lenders is fierce, and any technology or methods that produce better, more accurate, quick loans is highly prized as demand rises.
Balancing the equation between speed and accuracy is tricky to do, but it can be achieved by integrating artificial intelligence and machine learning into an already robust loan origination system, and making sure that your staff are on board with the new ways of working, and that they are excited about the potential of the technology.
The inefficient and slow nature of traditional mortgages has incentivized technology firms to develop agile mortgage solutions which have the ability to transform the lending process for bankers and loan originators.
Higher quality, faster loans are achieved by creating improvements at every stage of the process. There is no single aspect of the LOS that can be improved in isolation that will then go on to transform the whole process. It needs to be a systemic shift that creates efficiency in every aspect.
From quality onboarding to processing, underwriting, closing, and shipping—it’s all interconnected.
Begin at the Beginning
When mortgages are driven by technology, the engagement begins almost immediately. Initial documents are identified and classified using intelligent software within moments of them being received. A document indexing and filing robot handles almost all of the initial, time-consuming work of labeling, classifying and extracting data from documents, which is the stage where errors often creep into the process.
As the assorted loan documents enter the system, a cognitive robot should have the ability to:
- Perform automatic ingestion, classification, and delivery of documents to LOS folders
- Update the naming convention of specific documents to include the borrower(s) name
- Automatically create folders in the LOS for newly classified document types
AI Foundry  CEO Steve Butler explains the bedrock technology that underpins agile mortgages by saying, "User-guided machine learning, workflow, and business process management technologies are at the heart of our actionable intelligence solutions. We can take a document, extract intelligence from that document, and then start a workflow or robotic process automation instance to automate very manual processes.”
That ability has a dramatic effect on the speed and accuracy of any application. The automation and increased efficiency play out in a number of critical aspects of the loan origination system, each of which must be well-managed and efficient in order to produce fast, high-quality loans.
Onboarding: This is one of the most critical parts of the process. In order to obtain a mortgage loan, a borrower must complete an application and provide all required documentation to the lender, which must then be reviewed for completeness. At this stage, there is a massive amount of new data entering the system that needs to be extracted, stored, and checked. During this time, the cognitive robots review the application for completeness and perform a general eligibility audit on the initial application submission. Software that can measure the application against a pre-existing checklist and flag any defects early on is a vital part of onboarding and leads to higher levels of customer satisfaction and speed.
Managing Defects: Of course, there are always going to be issues in any loan application. With so many variables, document types, and the amount of information required to process a loan, mistakes will occur. But they cannot be allowed to languish in the system unnoticed. Cognitive robots are transforming the mundane tasks such as data capture, verification, and the transferring of information and slashing the time it takes and the cost of back-office processing.
According to ARMCO’s Mortgage QC Industry Trends Report  from 2019, “24% of all loan defects are attributed to missing, expired and/or incorrect documentation.” That’s the kind of information that really drags down the speed and quality of a loan. Defects must be flagged early and rectified in order to ease the flow of a loan through the system.
Collaboration: The days of teams sending documents to each other via email and waiting for a reply need to set firmly in the rear-view mirror. When the whole team is able to work off a shared CRM that shows the status of the application to everyone, it becomes much easier to speed up the loan through the system. You should not be held back by any one individual who happens to be absent, or unable to fulfill a certain step. By giving everyone on the team real-time access to the same versions of documents, it has the effect of eliminating confusion and streamlining communication. As a result, issues tend to be identified and resolved far more quickly than ever before.
Compliance: The quest for speed and efficiency cannot come at the cost of compliance. TRID regulations  have resulted in an increased need for oversight and, as a result, can lead to higher costs being built into the loan. That’s why the TRID Verification Robot is so valuable. It operates simultaneously alongside the other bots, operating as another safety layer. By focusing on documents such as change of circumstance, closing disclosure, loan application and estimate and the purchase contract, the TRID Verification Cognition Robot provides another invaluable layer that helps a loan hit the milestones that bring it one step closer to completion.
There is no doubt that automation is transforming the way lenders do business. In today’s market, where volatility and unpredictability seem to be ever-present, those challenges are being balanced out by a lender’s ability to provide fast, high-quality loans that take these risks into account and use technological solutions to manage them effectively.