Voice AI Is Changing How We Do Business, Starting With Loans

Forbes

5 days ago

Putting all the pieces together to secure a loan isn't easy. Back when I was a mortgage broker, much of my day was spent tracking down bank statements, W-2 pay stubs, and countless other pieces of paperwork. I kept all this in manila folders for my clients, alternating my time between sales prospecting and battling underwriters to get my loans through to closing.
The Loan Bottleneck Challenge
The many hours involved in pushing a mortgage across the finish line were daunting, the work tedious and dull, both for me and my clients. And if that's true for home loans, the problem is exponentially worse in commercial lending. Loqate, a global location data and address verification firm, cited a recent related statistic in its report from Forrester that '54% of people filling out financial application forms abandon them prior to completion, and on top of this, a recent study by SaleCycle saw financial abandonments reach an average of 76%.'
But here's what shocked me: for commercial loans, that number can jump even higher. All those abandoned loans represent entrepreneurs' dreams deferred, expansions delayed, equipment un-purchased, growth stalled. By the same token, they hint at money that's flowing into neither the pockets of loan officers nor the brokerages employing them.
An AI That Talks You Through the Loan Process?
Enter: Qualify.bot, a conversational AI meant to streamline this fractured—if not broken—process. Two young entrepreneurs, Yash Goenka and Rohan Datta, are behind the artificially intelligent platform. Their brainchild recently landed them in the Y Combinator accelerator program where they join the prestigious ranks of other startups, including Airbnb, Dropbox, Stripe, DoorDash, and more. That's also the same valuable network that a guy you may have heard named Sam Altman once ran as its president.
I had the chance to speak to Goenka and Datta about why Voice AI is upending how we do business, starting with loans. "We call abandoned applicants within minutes while they're still at their computer, still motivated," says Goenka. "Our AI will spend 45 minutes explaining what a debt service coverage ratio is without any frustration. Try finding a human who'll do that cheerfully on the hundredth call."
Instead of relying on a text bot with canned answers, their system automates the whole loan workflow. The power of talking AI agents assists borrowers from the moment they first fill out a form to when their completed application lands on an underwriter's desk. 'Think of it as your own 24/7 sales development rep, loan processor, and AI assistant, rolled into one,' Datta adds.
Why Voice AI is Different
Before delving into how Qualify.Bot works a bit more, let's discuss the phenomenon that is Voice AI. You've most likely had the annoying experience of dealing with what I will call 'dumb' automated phone agents. I even had one today that left me shouting into the receiver, 'Just connect me with a human!'
Voice AI is not that.
In May of this year, the Wall Street Journal described the next generation version this way: 'Automated voice programs are being upgraded from old-school systems with little or no AI to newer speech-to-text and text-to-speech models combined with large language models. If the technology lives up to its promise, the shift might improve the customer experience at a range of companies and reduce their costs in the process.'
"The convergence of three technologies changed everything," Datta explains. "Latency dropped below 500ms—fast enough for natural conversation. LLMs, like ChatGPT, can now maintain context across thirty-minute financial discussions. And voice synthesis became indistinguishable from humans."
While that's undoubtedly true, what's more important to this conversation is how voice AI stands to improve operations for tomorrow's companies, starting with lending. Returning to my own experiences as a loan officer, there were two major components to my job:
Voice AI, by way of Qualify.Bot, seeks to manage part one, handling the soul-sucking part of working in this business. Beyond pre-qualifying applicants and coordinating the paperwork needed to complete a loan, it also coaxes borrowers along, ensuring less of them prematurely exit the process.
If a borrower stalls in the midst of an application, for instance, because they cannot locate their EIN number—or because they simply find the whole thing too frustrating and convoluted to go on—the AI will step in. It will also happen in a more natural way than a text box pinging them with pre-programmed messages. Instead, a (natural-sounding) voice agent will call the borrower back to find out what happened and how to fix it. As you can expect, such handholding, if performed correctly, can lead to higher application completion rates and ultimately, more closed loans, a win-win scenario.
It makes sense that the lending industry is embracing such tech. Sophisticated AI seeks to alleviate the slog of paper-pushing to help businesses access the capital they need to grow. "Lending is uniquely perfect for voice AI," Goenka explains. "It's high-stakes enough that people want to talk to someone. Nobody wants to get a business loan through a chatbot. Yet it's also process-driven enough that AI can actually help. There are correct answers to questions like, 'What documents do I need?' AI can be trained to handle these perfectly."
It's important to note we're not talking about yesteryear's voice agents. Thanks to AI, they're getting better all the time. Much better. Anecdotally, I have interacted with Voice AIs that sound incredibly human, down to the nuances of person-to-person communication, even the 'ums' and 'uhs' indicating authentic-sounding exchanges. In fact, the only thing that alerted me I was talking to an AI was the unnatural pause it took between responses. Several seconds were often required, taking me out of the moment. This is but a slight hiccup that we can naturally assume will be corrected in short order.
The key part of this interaction Goenka and Datta have hit upon is the ease of voice versus text communication. Like it or not, we are no longer a literate nation. People don't read books. Sadly, they don't even read Forbes articles as much as they once did. Therefore, they have little patience with typing their way through complex commercial loan applications asking for financial statements, tax returns, and business projections. They just want someone—anyone—to solve their problem fast so they can get on to their next task.
From Loan Officer to Trusted Advisor
This represents the real utility behind Voice AI and the reason Silicon Valley, not to mention, Big Business, is betting big on it. Even so, the former loan officer in me can't help but wonder if this job will still exist in five years.
According to our young entrepreneurs, the answer is yes. As Goenka explains, 'Just like ATMs didn't get rid of bank tellers, Voice AI will not do away with the all-important human touch. In five years, AI will handle all application intake, document collection, and basic underwriting. Loan officers will shift into more strategic, high-value roles: becoming trusted advisors, closers, and relationship managers.'
A few years ago, I might've disagreed with him, so certain I was that AI would gobble up all jobs like the insatiable leviathan pundits warn us about. Not anymore. The fact is people want to do business with people. We like other humans. And we really like being more productive—any way we can. Voice AI is enabling such breakthroughs, leading to more money in the pockets of brokers, more deals for brokerages, more commercial loans, and more prosperity overall.
Now, that's a future worth celebrating.