Jorge Sun On How Using AI In Lending Benefits Bankers And Small Businesses
Applying for a loan is a pain, especially for small businesses. However, the era of technology has opened opportunities for automation and AI in lending. In this episode, Chad Burmeister is joined by Jorge Sun, the CEO and co-founder of LendingFront. This software application works with banks and leverages AI to loan money to small businesses more efficiently. Jorge explains why this process is better and benefits not just small business owners but bankers as well. He also shares some of his insights regarding the future of banking and where humans will play a role. Lend an ear and learn all about this innovative new platform.
Listen to the podcast here:
Jorge Sun On How Using AI In Lending Benefits Bankers And Small Businesses
I've got Jorge Sun and Adam. Adam with us as well on the marketing side. Jorge is the Cofounder and CEO of LendingFront.com. LendingFront is a software application that works with banks and other institutions, credit unions and leverages AI in its ability to loan small businesses money. We're going to dig into some of those topics but before we do that, we're going to get to know Jorge a little bit more as well. Jorge, welcome to the show. Thanks for being here.
Thank you for having me. It’s my pleasure.
You're in Cape Cod, Massachusetts. What's it like out there at this time?
It's really nice. That's why I'm out here. It's cooler than the city where we're based as you can imagine. I think you're down South. It's not 100 degrees here, it's more like 80 degrees so it's good.
We're in Colorado. We tend to be 70 degrees, 80 degrees. We did have a couple of 100-degree weather already this summer but that's okay. It's nice to have variety. The four seasons we have here for sure. For our audience to get to know you, I to ask the question about when you were younger like 6, 7, 8, 10, first memories to help connect the dots to what you're doing, what were you interested in then or what were you passionate about in the early days.
That was an interesting question I read and I was thinking about it. What I've always enjoyed was building or creating new things, whether it was putting a couple of toys together to build something different. I didn't go into art but I was a fairly decent artist using white paper to create new things was something that I always enjoyed. LendingFront is more of a physical representation of creating something out of my experiences. That's one of the things that I would say that resonates from back then to now.
If people are reading and they're finding themselves being a creator, cofounder CEO is sometimes a good position to exercise those talents. No doubt about it. What's the thread between what you're doing now, then how did you come up with the idea of the software that you deliver? How did you understand the gap in the market and then create it?
The ability to have something that helps you out of decisions is very important.
I followed the more traditional entrepreneur, which is you go to work, you get a job, not necessarily the job you want but you choose something that appeals to you. I got recruited to work with some of the best lenders in the country, American Express, JP Morgan, Capital One. I was always in a risk management data function, which requires a lot of creativity like how do you lend, when do you lend, what do you look at to lend? That's where I got insights into most infrastructures that exist either commercially focused or large institutions looking to borrow $10 million or very consumer-focused for us looking to borrow it, get a credit card.
There was a big gap in small business lending. They tend to straddle them both worlds a little bit. LendingFront came out of the first initiative being one of the early team members of a company called OnDeck where we pioneered the use of data analytics to lend. We were the lenders but using the internet as a distribution channel, acquisition channel, using different information like cashflow data. This is way back in ’07 to determine risk and completely different product structures, daily collections. That did well and went public but I, as an ex-banker always knew or always expected that those innovations should be in the hands of your standard device if we wanted to solve the capital problem, that small businesses face. What does LendingFront do? It’s taking the experience of having built a platform before, knowing how to lend to small businesses and now packaging an infrastructure that banks can use to lend more efficiently.
I remember 6, 7 years ago in downtown Denver, we were looking for space and OnDeck were upgrading their offices. I was the VP of sales for a tech company and we looked at the OnDeck office and it was so enormous. I was like, “How can you grow out of that space?” They did something right when it comes to the algorithms behind the scenes. I get it. There are so many lenders out there these days that come to, “We've got a good deal. It's very inexpensive to get capital from us.” You look at it and it's 22% or 34% a year in some cases. I'm like, “How do you define that as being good?” They're treating you as if your business is an extremely high risk, when if they could plug into QuickBooks and my Bank of America account and the basics, they could see, “You have three and a half years of history. You've never missed a payment.” Know exactly how much they could lend. I assume those are the kinds of things that you help banks do is analyze those kinds of risk profiles and make good decisions on lending.
It's enabling them to have access to those new data sources and connections because most of our clients are on legacy systems. They're not designed to leverage that infrastructure. Their hands tend to be tied.
How does AI work? I'm assuming there are APIs into those feeds and it can all happen in minutes instead of hours, days, weeks in some cases. What's the black box look like on the back end? How does that work?
It's a little different for each client because again, the way they originated, the way that lens a little different but the basics are the same, which is how do you make a banker more efficient? That's by taking away all the basic, very clear, easy to a decision, applications, processes, you name it. AI is designed so you can get rid of all the potentially rote steps that a lot of banks forced her bankers to do. That artificial intelligence picks up quickly and you'll notice this very quickly too is I always ask our clients. If you have a client that's been with you for five years, they keep whatever $25,000 in the bank, their owner has prime credit business, has five employees and generates over a million dollars in revenue and all they want is $25,000. Are you going to send that to a banker to analyze and ask for three years of financials, etc.? Are you going to let the machine figure that out for you? A lot of those things tend to pop up. The most obvious ones are the fast noes. Also, one of the things banks are good at, in a bad way, is holding you hostage for a couple of weeks to tell you no.
I had a guy, a friend of mine move out of state to Houston. They were three 3, 4 weeks in. At first they said, “No problem. You're an ex-CMO of a big company. You've got a lot of money. You've got everything. You're fine.” The day before they were supposed to move, the trucks were showing up, “Sorry, you don't qualify anymore.” He was like, “What are you talking about?” That can cause some hiccups, it would have been better to know at day two that it was a no.
AI In Lending: Are you really going to send that to a banker to analyze and ask for three years of financials, or are you going to let the machine figure that out for you?
It all stems from the fact that a lot of the decision is left to individuals. There are some guidelines but those guidelines are written and most banks let bankers try to remember what their credit policy is in their head. As you can imagine, sometimes they forget or they get bumped up and somebody goes in a bad mood. Those are the applications. The other big application though is also trying to figure out not just on the risk side how to lend but also what can you predict when somebody will need capital by looking at patterns of their cashflow? If it's too choppy. Consistent but choppy, it's a great time to go out and say, “Do you need a line of credit?” You're growing rapidly and you need to expand. A lot of signaling to make the outreach a lot more efficient.
Does your software automate those steps as well to send the email to the banking customer and say, “We noticed you're choppy? Let's get you some cash.”
What we provided is the infrastructure to automate that process. It is still because the bank is the lender of record. They are the ones who have to decide the base level triggers or accept the recommendations to say, “Send it to a banker, call it the customer or send an email.”
I had a colleague of mine on the news. They asked him about what do you see going on with small businesses, consumer confidence and growth. He haphazardly pulled it from the sky and said, “I think it's good. Consumer’s confidence is at a high.” Do you have any information based on the data you're seeing of loans? Are there any bumpy waters ahead or are we in pretty good time?
Most of our clients lend to small businesses. They usually only lend based on the strength of that business. We saw them stop lending during COVID. They stopped, you see this drop but they're lending at a higher rate than pre-COVID. That tells us that there's a lot of confidence because the lending is not based on last year's tax statements for two years your collateral value, which is meaningless. If you think about the last year, that doesn't mean anything. They're lending based on the amount of money that they're processing, which means people are spending money in that business or store. If that's growing, it’s giving a lot of our clients confidence to lend and it's a 45-degree line to the right.
That's good to know. That's real-time data. The next question is if you didn't have AI, would your company even exist? How would it exist without AI?
The difference is the amount of their amount of services or the willingness of some of our clients to be excited about the platform because a big chunk of what we do is try to help them automate processes, decision points. It would be up to them to have an AI solution and/or leverage people to do it. That defeats the purpose. Having the ability to have something that helps you out of decisions is very important.
You’re always going to need a banker because you still need to know who the customer is.
We have an email technology through a partner that's an Israeli company and it's white-labeled. Like you provide services to banks that white-labeled their banking app, ours, is a white-label someone else's product. They monitored our manual responses to emails over the last 6 to 12 months because we set it so that it was a pretty low threshold. We didn't want to mess anything up. They monitored us and they said, “Chad, you're at a 98.9% satisfactory rate where you can turn over the keys to the car to the AI and let it respond to the prospects.” There were 2 or 3 minor tweaks because for the 1% if you responded and said, “It sounds you're like not interested.” Meanwhile, they said, “Let's meet,” but the AI got it wrong then that would be a problem.
There were ways to program to solve for the 1% but now I don't have to go in once or twice a week and respond to 39 and read emails. The AI's doing it. Now I can go in and look at 1 or 2, the exceptions versus the norm. That's what makes sense to me, a $25,000 loan or a fast no. These are the things that AI can do. Across large organizations, the elephant in the room for a lot of companies is, “Will this mean that I need less?” If you have a thousand people on staff, you might be able to get by with less. Most companies don't want to get by with less, they want to grow and they're going to put their money back in. If they save a little bit here then they'll buy a new building, equipment or software.
The big reveal there is most banks have a number of bankers, they're expensive. They tend to be overworked to a certain degree. They tend to ignore small businesses because they spend the same amount of time on a $25,000 loan as $1 million. The reason you see so many small businesses not get help from the bank is the banker has to make a choice. Do I spend it looking in servicing a $1 million request or that $25,000 request? What this type of technology and AI does allows the bank to service more customers. It makes their staff more efficient because now instead of getting bothered by these small requests, they can go focused on the larger ones, be a lot smarter and not reduce staff. If anything, we've seen clients add staff to your point because they're being so successful that they're like, “We need more support”
Now they can go after a new market and say, “Let's open it up to small businesses.” They can afford to hire a new resource or two. That's awesome. The last question of the day is where's it all heading 3, 5 years from now? Does it replace the human at some point? Are there always human eyes required on these?
There are always human eyes, at least for now until the infrastructure around how businesses form, how you validate a business until there is no more cash. You're always going to need a banker because you still need to know who the customer is. You need to understand what they do to some degree because fraud is an issue. It may not be a fraud but borrowing because you're not doing as well as you should. It’s still an issue but things will be significantly more automated and self-preserved. Where the banker's going to add value, the person is more on the tailored requests. I'm packaging a relationship versus one-off requests. Now that I can validate you especially if you belong to some platform like your Amazon seller. I know you're real, you have a lot of customers looking at your products. I can track you over time. If I can look that at information, I can make decisions relatively quickly and don't need anyone. If you want to buy a building until all the infrastructure to support buying a building and creating a lien until that becomes digital, you're still going to need some people.
Anything else you'd like to add?
No, we are at the beginning of lending the use of AI. This is the beginning. Finally, things are catching up in terms of supporting and feeding the data you need to make things a lot more accurate and efficient.
AI In Lending: Bankers spend the same amount of time on a $25,000 loan as a million-dollar loan. So, many small businesses don’t get help because the banker has to make that choice.
Is it a proprietary AI system that you've built? Are you finding some of these tools out there like Watson, Google and Microsoft? They're all coming up with the body of AI. Do you have to build your own? Can you attach to something already available?
We use whatever's available. We're not doing crazy stuff or like, “Do you give this guy a loan or not?” That is not rocket science, it’s on basic things you look at. We don't need to reinvent the wheel.
That's where things are headed as standardizations occur. The flywheel's already built. You might as well leverage all that body of work of 100 of man-years that’s gone into it. It makes sense. Awesome conversation, Jorge Sun, Cofounder CEO of LendingFront.com. If you know anyone in banking, if you are in banking, highly encourage it. If you want to expand your footprint to reach out and help support small businesses, this is a cool software technology that can help you do that more effectively.
Thank you so much, Chad. I appreciate it.
Thank you for being here. We'll catch you on the next show. Cheers.
Important Links:
About Jorge Sun
CEO & Co-Founder
Founding team member and Chief Credit Officer for OnDeck. Most recently, Head of Small Business Credit at Capital One Bank.
LendingFront’s revolutionary platform dramatically reduces the cost and significantly simplifies the processing, pricing, loan management and credit monitoring of business loans. This enables traditional lenders to rapidly scale into the online lending market at a fraction of the time and money it would take to build or modify existing systems.
