Log in     Support     Status

How to Solve Your Biggest Mortgage Lender Origination Problems

by | Jan 4, 2021

Dealing with mortgage origination as a mortgage lender was once a time-consuming process because of the high volume of loans as well as the time taken to process them. Working with underwriters is already complicated and can lead to extra manual work that sometimes leads to embarrassing errors.

Using manual processes in mortgage origination is no longer a recommended practice. Numerous things can go wrong, leading to critical mistakes and potentially creating a lack of trust.

An ideal solution to these problems is to automate many repetitive tasks. What kind of platform can you use to make this happen? Let us show you with our mortgage automation platform, Capacity.

Automate the high volume of mortgages

You likely have a huge number of mortgages to deal with and not enough time to process them all. Fortunately, automation helps on this front thanks to Capacity’s use of artificial intelligence to speed up many tasks.

At the head of this is automated onboarding support for new sales and operations personnel. This already saves time on training when a large pile of mortgages is waiting to be processed.

Most useful is the ability to use Capacity’s AI capabilities to have intelligent conversations with new leads. Facilitating the loan process through the use of artificially intelligent chatbots is a revolutionary way to speed up mortgage origination like never before.

Above this, though, is the ability to streamline all manual processes for mortgage lenders and use automation to eliminate all repetitive work.

Allow automation to handle manual tasks

For years, mortgage origination involved going through numerous steps to complete the cycle. It starts with pre-qualification, the application process, underwriting, credit decisions, quality control, and loan funding.

All of these steps involved manual tasks for decades, even in the age of computers. Some of those tasks simply had to be done manually with the thought it would avoid mistakes. Underwriting, for one thing, usually involved manual research by mortgage lenders to make sure a client was truly eligible for a mortgage.

Capacity changes this for the better by bringing automation to all these steps. This includes using an AI-powered chatbot (internally with mortgage brokers and externally with borrowers) so mortgages can get approved in half the time.

The technology involved here is NLP or Natural Language Processing. This amazing tech allows your chatbot to become more intelligent by processing questions and bringing accurate answers from your knowledge base directly to both mortgage brokers and borrowers. It gives the capability for lenders and borrowers to get instant answers from the chatbot as if they are talking to a co-worker or friend. 

Eliminate human error

All those manual processes lenders had to use for years led to inevitable mistakes. A lot of banks still use a paper-based system in originating mortgages. This leads to more errors in fact-checking, not including possible mistakes while researching the borrower’s background.

Automation and machine learning are changing the game at Capacity. In turn, it also brings more engagement with the client while the automated system researches and calculates monthly mortgage payments.

A mortgage automation platform can authenticate data, pre-fill out forms, monitor markets, and verify credit scores. Plus, AI helps make stunning predictions about when someone is going to buy a home, making the timing and approach of clients better targeted.

Automation also improves the underwriting process. It reduces the average time-frame of 47 days for a mortgage down to significantly less. Eliminating optical character recognition (OCR) is additionally possible using Capacity’s automation system. Those who’ve worked with OCR know it’s not nearly as accurate as automation is in interpreting client data.

Reduce the time it takes to process mortgages

As noted, the length of time it takes to take mortgage origination from pre-qualification to funding is sometimes close to two months. Other times, it can take longer, depending on the client research involved.

The big challenge with that is clients expect faster turnaround times from what’s typically common. Using automation to cut this time down to weeks rather than months helps improve client trust with your company.

One part of the mortgage process that’s often overlooked (and involves a significant amount of work) is the closing. When a mortgage is about to close, it usually means considerable paperwork, which means manual work prone to more errors. Time saved with pre-filled forms shaves off a couple of weeks of this exacting manual work.

How are mortgage companies changing thanks to capacity?

It’s worth noting some major mortgage companies managing to change their lending processes using Capacity for mortgage origination purposes. Paramount Residential Mortgage Group Inc. (PRMG) used our machine learning to provide more solid answers for customers about lending. PRMG also used automation to help provide better support for their loan originators and employees.

West Community Credit Union (WCCU) increased total asset holdings by 50% using our automated tech. They used it to help inform their lenders through the chatbot. Their philosophy is that the more educated the client is, the more confident they are in wanting to go forward with a mortgage.

AmeriSave used Capacity to help differentiate themselves from others like them. They created a more personalized approach to answering questions for their borrowers.

Bring humans into the loop

Let us reiterate humans are still important in speeding up mortgage origination. We’re subsequently helping their workload, but also bringing them into the fray when the automated program can’t answer a particular question.

In this instance, we use Human in the Loop technology (HITL) that automatically routes a client to a human being when necessary. Sometimes it does take a real human to answer a unique question related to lending.