The Technology Podcast: Generative AI Takes Center Stage


Artificial intelligence is revolutionizing every aspect of the mortgage industry and as companies create systems that can help from origination to servicing, leaders speak highly of the possibilities ahead when it comes to accuracy, efficiency, and enhanced customer service.

Listen to our podcast on the subject:

At the beginning of the year, leaders at Rocket Pro TPO announced a new AI-powered hotline for brokers. Executive Vice President Mike Fawaz told brokers they can expect more tools to be released over the course of the year.

“Know that in 2024, AI and technology and really putting tools at your fingertips is our mission and these are the things we want to be able to deliver to you every single day,” Fawaz said.

“And we’re going to show you how we can take AI from a world where it’s just like out there and it’s fantasy, and it’s this buzzword, and really show you what we can do.

Of all of the lenders pushing advances with AI, Rocket Companies stands out.

CEO Varun Krishna is laser-focused on the topic, maintaining that new technologies will rewrite the rules of the game and fundamentally change how the mortgage industry operates.

He has told investors that they expect to see better client experiences and capacity management, saying during an earnings call after the results for 2023 were announced that AI lets computers do what they do best while allowing team members to focus on what they do best, which is fostering relationships with clients.

Since Krishna started with the company in September, they have introduced Pathfinder, an AI and machine learning-powered search engine for mortgage bankers, brokers, and underwriters.

They have also talked about their technology for calls handled by Rocket’s client experience teams, including loans from mortgage brokers who work with Rocket Pro TPO.

We spoke with Mortgage Loan Originator and CEO of Agave Home Loans Marshall Gottlieb.

The company was listed as a 2023 top performing partner for wholesale and correspondent lending by Rocket Pro TPO. Gottlieb said Pathfinder has simplified the search process for his team.

“When you go into Pathfinder and type in, ‘Get more income out of a self-employed client,’ it’s going to give you a list of all the things that you are allowed to do and what questions you need to ask,” Gottlieb said.

“A lot of lenders, what they have is a PDF, and you can Control Find inside this PDF to go read about self-employment income. But that’s a lot of information. There could be 20 pages about self-employment income.”

Pathfinder has also helped them expedite the documents realtors need when they are helping buyers make offers.

“You can turn around a really strong pre-approval within 30 minutes as opposed to waiting a day or two days to find out from an underwriter if the person does or does not qualify for a new mortgage,” Gottlieb said.

In May, The Mortgage Note sat down with Shawn Malhotra and Dan Vasquez to learn more about how Rocket is harnessing the power of AI.

Malhotra had just been hired as the first-ever group chief technology officer for Rocket Companies. Vasquez is the AI strategy lead for Rocket Mortgage and started as a software engineer in servicing in 2013.

Malhotra said that Generative AI will be so transformative that it will differentiate this generation from others.

When asked how it will affect the housing market, Malhotra said large language models and Generative AI are capable of much more than the technology that is currently being used.

Malhotra explained that new systems have an immense amount of knowledge when you get them off the shelf. It’s general knowledge that companies are refining with data and human expertise.

“You’re using humans and data to improve the system. So if you have got human expertise and you’ve got lots of data for a specific domain, you can take these powerful off-the-shelf tools and point them toward very difficult-to-solve problems and have it help you do so,” Malhotra said.

Malhotra said the mortgage industry is a good match for Generative AI based on the tasks that need to be performed as part of the homeownership journey.

“What’s exciting to me about the homeownership journey is that whether it’s finding a home, whether it’s figuring out how to get a mortgage for that home, whether it’s servicing that mortgage, many of those jobs to be done, those tasks that we have to do for our clients, are right in the sweet spot for the kind of automation that Generative AI enables,” Malhotra said.

In April, company leaders announced that their patented artificial intelligence driven technology platform, Rocket Logic, had saved more than 5,000 hours of manual work for underwriters in the month of February.

This platform makes the mortgage process more efficient by combining insights from more than 10 petabytes of proprietary data and 50 million annual call transcripts with state-of-the-art deep learning and Generative AI systems, allowing Rocket Mortgage to close more loans faster, according to a press release.

Vasquez talked more about it.

“It’s a platform for marrying the power of AI, the automation that we’re able to create with new technology, and our data platform, bringing all of those things together and allowing us to just create the absolute most efficient mortgage experience not just for our team members, making our team members more efficient, more scalable, more accurate, we’re able to then pass on those benefits to the client. So the client then is experiencing a mortgage process that feels a lot simpler, a lot quicker, a lot more accurate. We’re able to ask the right question at the right time, in the right way,” Vasquez said.

“All of this helps create a more natural, more efficient, more accurate application experience.”

The company has also unveiled Rocket Logic – Synopsis, the next building block in the Rocket Logic ecosystem. It can listen to calls and point mortgage professionals in the right direction in real-time.

“If I’m talking to you and this is our first interaction, the tool is going to help me better understand you. It’s not going to replace the conversation but it’s going to help me better understand what your unique needs are. It’s going to be trained on my data and say of all the different products or all the different ways I have of helping you given this conversation right now,” Malhotra said.

Rocket Logics – Synopsis was built on Amazon Web Services and uses Amazon Bedrock. The tool also supports the servicing team that oversees Rocket Mortgage’s 2.5 million serviced loans.

It is estimated that 60,000 hours will be saved annually with the technology, which leverages AI to listen to, transcribe, and search client calls.

Vasquez predicts that because of new technologies the future is going to be delightfully simple.

“What GenAI is letting us do is really get back to, again, that kind of human level of communication and bringing the experience, bringing the entire mortgage process down to the idea of, ‘Can I just fire off a text or email?’ or some simple human-like communication just knowing on the other side of that, whether it’s a human team member, whether it’s an automated process that I’m going to be met instantly with a response that is personable, that gets right to the point, and that solves my problem,” Vasquez said.

Vasquez said working with AI is exciting because it offers a renewed sense of wonderment for people who work with technologists.

“As a technologist, you used to be able to have really fun conversations with folks. You used to be able to really wow them with, ‘Oh wow, you can do that? That’s possible? We can build that experience?’ And it’s fun and again as a technologist, those are the kind of conversations you want to have and that is the kind of stuff, that’s why you get into building software, or at least that’s why I did,” Vasquez said.

“With AI now, I really find we are back in that mode of, ‘Wow, I had no idea that was possible.’”

Rocket isn’t the only company working to stay ahead of the curve when it comes to making the best use of AI.

Other players in the industry are finding ways for technology to streamline their businesses and enhance client experiences.

Toby Wells is the president of Cornerstone Servicing in Colorado. They are taking advantage of AI using virtual assistants and he thinks technology is going to play a large role in servicing mortgages moving forward.

“We’re looking at using it, and I think other servicers use it, in data and analytics, things like when you’re transferring customers, there’s lots of data and information that have to be used and mapped. One servicer to another, it might be the same information but we call it something and somebody else calls it something else. AI can help you actually map that more efficiently and more effectively.”

Other uses include projecting the possibility of customers having a hardship and guiding employees through phone calls with clients.

Wells said that AI can be used to train agents, teaching them how to approach difficult conversations.

“For example, it doesn’t want you to say, ‘I cannot do something,’ or ‘No.’ But it likes it if you say, ‘I can do X if you provide this information.’ So, it’s conditional. And frankly, it’s a wonderful coaching tool,” Wells said.

Blue Sage Solutions in New Jersey provides technology to the mortgage industry.

They offer a lending and a servicing platform, working with a variety of clients, from credit unions to independent mortgage banks and top 20 lenders.

Chief Operating Officer David Aach spoke with us about how automation and AI are influencing the tools they are working on.

“One of the ones that we’re pretty excited about is a tool to do income analysis where it’s basically accepting all the documents, paystubs, W2s, tax returns, and doing a lot of the work that an underwriter today does pretty much the way they have been doing it for the last 20 years to calculate how much income does a person really have that I can use to get them qualified for this particular loan, so we’re automating a lot of that using AI kind of technology,” Aach said.

Aach said their products help underwriters be more productive.

“If you can pre-underwrite a loan using AI technology to do income analysis, and allow an underwriter to basically take a set of results that have already been kind of cleaned, and you have looked at all the documents, and you’ve looked at all the data, and you’ve put it all together in a nice, neat package, and say, ‘Okay, do you want to approve this?’ that saves the underwriter a lot of grunt work,” Aach said.

“Maybe the old-fashioned way, an underwriter could only do three loans a day. Using these kinds of tools, they could maybe do six or ten. That’s a big benefit to lenders.”

When asked what the response has been from people who work with their systems on a daily basis, Aach said it has been good.

“The worst job on the planet is the person who has to do stare and compare, you know? Or look at a document, take data that they see on a document, and type it into another system,” Aach said.

Aach said AI also improves accuracy by alerting lenders to anomalies that humans may not catch.

Tell us what you are seeing in your market. Email Editor Kimberley Haas at [email protected].

Read More Articles:

Rocket Team Members Explore GenAI During HackWeek

NAR’s $418 Million Settlement: Where Are We Now?

Sports Venues New Stars In Real Estate Development

Sign up for our free newsletter.