How Artificial Intelligence is Changing Consumer Lending with Jeff Keltner

Intro: Helping community bankers grow themselves, their team, and their profits. This is the Community Bank podcast.

Caleb Stevens: Well, hey everybody and welcome to the community bank podcast. My name’s Caleb, and I hope our time together helps you grow yourself, your team, and your bank’s profits. Today’s guest is Jeff Ketner, he’s an executive with a FinTech called Upstart and our discussion today is all about AI artificial intelligence and how it’s shaping the banking industry specifically as it relates to credit scores and consumer lending. Jeff’s a former leader at Google and we talk all about the lessons he learned while working for one of the world’s most innovative tech companies, and he’s got a podcast himself called Leaders and Lending, he is an overall great guy and so I think you’ll enjoy this discussion. Before we get there, as mentioned before on this show, you probably know very well by now that rates are likely on the rise this year and one of the ways we’re helping banks like yours take advantage of this rising rate environment is through the ARC Program. The ARC Program allows you to offer your borrowers up to 20-year fixed-rate loan while you earn a floating rate on your balance sheet, as well as additional upfront fee income. The best part is this is not your typical interest rate swap program, there’s no derivative on your book, there’s no hedge accounting, no dot frank reporting and so not only will your lenders love it, your borrowers love it, but your CFO will too. To learn more and get started, go to southstatecorrespondent.com/arc. Hey, we’ve got a number of resources in the works on how to price more profitable loans just in general coming soon and so stay tuned to the podcast because, in a couple of weeks, we’ve got some really cool free resources for you and your teams on how to price more profitable loans in 2022. With that, here’s my conversation with Jeff. Jeff, welcome to the podcast.

Jeff Keltner: Thanks for having Caleb, I appreciate it.

Caleb Stevens: So, you’re out in Pasadena, California, which I’m really jealous of. We talked before we started recording here that the last time, I was out there was for the Georgia Oklahoma rose bowl, which was one of the best football games I’ve ever gotten to see. So, I love where you’re at and I’m jealous that you get to experience Pasadena every day.

Jeff Keltner: It is a great place to watch a football game, I’ll tell you it was 32 degrees here this morning. So don’t be too jealous today was not a typical South Cal, sunny and warm day, but it’s usually pretty nice.

Caleb Stevens: That sounds like the weather here in Atlanta, it’ll be 32 in the morning and up close to 75 in the afternoon or 80, and you’re burning up, so unpredictable. Well, give us a little bit of your backstory in terms of how you got into the FinTech world, the banking world, you’re a Stanford grad, you spent a lot of time as a leader at Google. Give us sort of the backstory of how you kind of came to do what you’re doing now, serving banks.

Jeff Keltner: Yeah, I mean my backstory, I got to Stanford as I thought I’d be a physics major and I think like so many people at Stanford do end up studying computer science and computer engineering and kind of cut my teeth. By my senior year, I’d realized I was not the strongest computer scientist, maybe that’s because when you’re at Stanford, there are so many just incredible computer scientists, you don’t quite feel like you measure up. So, I kind of decided to put my career at the intersection of business and technology, not being an actual developer, but being kind of on the business side of those spaces. I spent my first couple years at IBM and sales of which you might call a mainframe computer, so old school technology, and then I landed at Google right at the moment when I was kind of hired to start what is now Google cloud and kind of G suite or Google workspace, the Gmail for businesses. So, it’s kind of the first business hire into that product, which obviously was kind of a catching a tiger by the tail kind of experience and ran the education group there for a number of years and then the kind of enterprise devices group with Chromebooks and stuff and then the head of the at division Dave Gerard left to found upstart and he said, “Hey, this has been a pretty good ride, like what’s next”, and I think the thing that was really exciting to me about upstart was what you realize in a place like Google is that the technology is really interesting, but a lot of the value really is created when you apply the technologies to transform industries. It’s not the technology for the sake of tech technology, it’s what can we do to make X, Y, Z better by leveraging these technologies and the opportunity to improve the financial services industry by applying more modern technology, particularly things like machine learning just felt huge and scope and so immediate and capability that it was really exciting to me. And that’s kind of how we ended up deciding to come and do this.

Caleb Stevens: What were some of the biggest lessons you learned working at Google when you started in 2006, was it?

Jeff Keltner: Started 2006, yeah. You know it was interesting, I learned kind of the standard. I guess I lived the standard crossing the chasm experience, I mean we were selling cloud computing in 2006, and cloud computing wasn’t a thing. I think Eric Schmidt might have coined the term ‘cloud’ back at that point and I remember one CIO from a university we were selling to, and they said, I think anybody who outsources their servers will be fired for gross incompetence in the next three years. Well, I think anybody who doesn’t will be fired for gross and competence, so I kind of learned what it’s like to be a pre-product market fit when the market’s not where you’re at the same time over the course of four years, we went from no product to 70% of the universities in the United States using that product.

So, I think I had the experience of what it’s like to cross the chasm in Jeffrey Morris’s terms and go from beating your head against a wall to really just trying to keep up with the demand for what you’re doing, which was a really interesting experience. And I’d say the other thing was the comment I made about, we started to see in schools, the value that was unlocked, not by the technology, but by when you could enable collaboration for instance, how they could change the teaching and learning experience in the school and that reality of what technology can do in context was really powerful to see.

Caleb Stevens: We had Joe Oman be on the podcast about six months ago, and he’s the former CEO of Sea World and a number of theme parks and sob. He actually was part of a startup years ago, back in the late nineties, sort of a.com kind of thing and they were selling cars on amazon.com. Jeff Bezos was on their board and he kind of looks back and he says, I think we were almost too ahead of our time. He says you can be too late to the party, or in some cases you could be almost too ahead of your time, it almost sounds like the kind of early on you maybe had that experience but eventually, you’re able to sort of cross that chasm as you say.

Jeff Keltner: I think some businesses cross the chasm and are successful. There are definitely ideas that were there and just, weren’t quite the right time for the market’s point of view from a technology point of view, couldn’t quite make it work. You see both, but the experience of in the dramatic shift in what it takes to operate a business when you go from trying to find that product-market fit to, ‘Hey, we’ve got something really exciting here growing.’ I mean, everything about the way you operate that business shifts at that moment, and I think something you can only really appreciate by having gone through because it’s kind of a dramatic change in the way the business looks and feels.

Caleb Stevens: One of the things I’ve admired about Google and you were there, so correct me if I have this stat wrong, but I heard at some point, Google realized that culture best develops in groups of 150 or less. And that’s about the number of real relationships you can probably maintain at one time and so they would intentionally sort of strategically make sure their departments and their different office spaces had no more than 150 or so in the building. and so, culture could develop. Was that your experience and what did they do from a cultural standpoint? Cause I’ve always sort of seen them as kind of on the cutting edge, not just technology-wise, but also workspaces and people and HR and culture and that kind of thing.

Jeff Keltner: I don’t know if I ever saw that, I’ve definitely seen that research and they were cognizant of that, but I don’t know that I ever saw that directly impacted in my role. It was a fascinating place to work culturally, we were still a very office-oriented culture, but the campus was so big that sometimes you had to drive 10 minutes between meetings that were back-to-back, and so it could be pretty tough. But I think the core thing I took away from Google’s culture was the speed at which they tried to operate and make decisions and move forward and accepting that you weren’t going to have the right answer upfront, but that if you got into the market and learned from your experience, that was going be the way you found the right answer. I think they’ve not always executed that vision well, um, and some of their products, but that certainly was what we found and what we saw. There’s this very interesting balance in of kind of two cultural norms I think in Silicon Valley, one is the kind of Steve jobs you’re going to like what I tell you; I think it often kind of exemplified by the Henry Ford quote if I ask my customer what they wanted, they would’ve asked for a faster horse, then they kind of put the customer at the center of everything you do and make sure the customer is your focus. These two things are sort of intention, but not really, I think they’re really a focus on understanding the customer’s problem and iterating your way to the best solution, not necessarily the one your customer knows that they want, but the one that is really solving their underlying challenge. I mean, in the Henry Ford quote, it’s not like you were ignoring the fact that they wanted to move faster, you just weren’t giving it to them the way they expected and I think that balance is really interesting, something Google led its best has done pretty well.

Caleb Stevens: Well, one of the things that struck me about your story is it sounds like you had a mentor, somebody you looked up to that then said, “Hey, I’m going to take what I’ve learned, go into a new venture and open an up an opportunity for you and probably a lot of other folks.” Talk about sort of the value you for maybe the younger leaders listening that maybe don’t have sort of a mentor or somebody in their life that they can really learn from, every leader I’ve talked to across the board, had what I call just quote a person that invested them, took a risk on them, opened up a door that maybe they weren’t even ready for themselves that believed in them and, that person took advantage of that opportunity. Would you say that’s sort of the case for you?

Jeff Keltner: For sure, when I used to say the CEO of Upstart’s a guy named Dave Gerard and he was the head of Google enterprise when I hired there and he hired me. I used to say he hired me into Google and then he hired me out and my mother-in-law thanked him for one of the two, I think I’ve got her around on the second one, at this point a decade in, but Dave was really the thing I think that was amazing is what Dave has done for me throughout my career and what a good mentor will do it. But frankly, just my advice to anybody would be to look for are opportunities that feel a little outside your stretch. I remember when I got that job and I said, well, what do you need me to do? He said I need you to go out and make this product successful. I said, well, how do I do that I’m 26, I don’t know how to do that. He’s like figure it out and it was kind of that slightly scared version of like, I don’t quite know that I am capable of what you want. I feel like I could probably figure it out, but it doesn’t feel like I have the qualifications. I think some people have described this as people who are punching above their weight class and I think you always want to find those opportunities to punch above your weight class and a good mentor will help you find them and help open those doors for you where you go. I know they say you need five years’ experience and you’ve got two, but I think you can handle it and you kind of go, I better justify that choice, I better make this person look good. So, I think Dave has done that for me at multiple points along with my career but I also think just from when I look at career opportunities or advise people how to look at them, if you feel like, yeah, I know how to do that, for sure. You’re probably not reaching enough, it’s not going to stretch you and push you to new capabilities, new opportunities and so that’s what I’m always looking for, like where’s the thing that feels a little scary, maybe a little outside my comfort zone feels a little outside my capability, but somebody else believes that you can do it. That that sometimes is what you need.

Caleb Stevens: And it sounds like you’re out of your comfort zone, but you’re still within maybe your strength zone. So, it’s part of your capabilities, you’ve got all the capability to do it, but it is a stretch. I think of the balance there is, is probably a key.

Jeff Keltner: It’s funny, you don’t want to jump off the deep end, but I learned this when I was an athlete in high school, which is a long time ago, but I think one of the most valuable lessons you learn is that you’re always capable of a little more than you think. A good coach will always push you and you go, I can’t quite do you, you can gimme five more, you can do one more, whatever it to is. You always have a little bit more than you think, the same is true in your career when it feels a little bit outside your reach to you, it’s probably not right. But you feel like it is because you’re at the edge of your capability and it feels uncomfortable. You don’t want to go there and find, I have no idea how to do this, don’t go there.
But usually, the thing that feels a little outside your comfort zone is actually right where you need to be and it just doesn’t feel safe, it doesn’t feel comfortable because it’s not a guarantee like you’re pushing the limits. And that’s what that feels like, I think a good mentor is one that helps you understand when your capabilities are really there, even though you don’t feel the confidence is there, that you can and helps you find those opportunities that are going to push you a little farther than you thought were possible, but really are what you’re capable of and that’s how you expand your capabilities and how you expand your limits.

Caleb Stevens: Well, give us a quick flyover of upstart and all the things’ you guys are doing to serve the financial industry.

Jeff Keltner: So, I will try and do that briefly, but the core insight, as I said that we came to is that the financial services markets were inefficient, that there was an opportunity to do better by primarily in our mind consumers. And the core thing we were looking at was the ability to identify creditworthiness among consumers. One of the first pieces of research we did with TransUnion showed us that more than 80% of the American public had never defaulted on a personal credit obligation, but less than half of the American public had a credit score that would qualify them for credit from a traditional bank. And you kind of went well, there’s a huge gap here, many more people will pay back a loan than we would give the loan to. Isn’t there a way that we could figure out who some of those people are and make credit more accessible and more affordable for a broader set of the public and so we really started down the path of applying more sophisticated technologies. Today, I would say machine learning, or what many people would term artificial intelligence to the problem of identifying creditworthiness and individuals, and we kind of along the way determined that not only were we not identifying as many creditworthy people as they are, but we were putting a lot more hurdles in the way, people in terms of how we verified identity or income. And so, there’d be all these processes that put friction into the process of borrowing money that could be removed with the smart application of these kinds of automation and artificial intelligence technologies. So, we really work with banks and credit unions to take the capabilities of AI and machine learning and apply them to consumer credit today, primarily in helping them kind of launch and run personal unsecured lending products and auto refinance lending programs.

Caleb Stevens: Well, this is probably me just being a little bit uneducated on the history here, but do you have any sort of sense of what the backstory is to credit scores in general? I mean, I would think back in the day there was maybe more of a manual underwriting process. I would think credit scores drastically speed up than simplify the process and to your point, maybe have some unintended consequences, but what’s sort of the backdoor story for credit scores in general, because you hear people joke all the time that I could have a million dollars in my savings account, but if I don’t have any credit, if I don’t use a credit card, I probably can’t get approved for a loan even though I could easily qualify if you think from more of a practical standpoint.

Jeff Keltner: Yeah, and you’re totally right. I mean, credit scores came around and I think the eighties were kind of revolutionary, if you think about it like we had been totally manual, like you had looking at somebody’s personal balance sheet, sitting across a table, am I going approve Jeff for a loan? Yes, or no, kind of an individual decision and so credit scores were the first time we had some sort of a universal algorithmic look at a person’s credit history and could rate it. I just think the challenge has been a couple of fold, one credit score has not kept up with the times, they haven’t kind of moved the state of the art forward that far. And secondly, the concept of a three-digit score, that’s going to encapsulate the creditworthiness for someone on a $2,000 loan, a $500 credit card, and a $2 million mortgage, those things just don’t really make sense. So, I think that the score is a concept that is really hard to wrap your head on how it can be inclusive for all kinds of lending because you’re not kind of an average risk for a huge mortgage and a small loan.
You could be a really good risk for one, a really bad risk for the other and when we reduce you to a score, we can’t capture that. And the other part is that scores are inherently rearward looking, they’re based on the history of your credit. So particularly if you’re young, you come out of college, come out of high school, you get a job, you have no credit history. So, does that make you a bad risk? I mean it does if all I’m doing looking historically, but practically there are pieces of information we know about you, are you employed? How much do you make? What kind of industry are you in? Maybe tells me something about your likelihood of unemployment. That could be additional factors. You can look at to understand the risk of some one, but just aren’t part of a credit score. And so, I think there’s a lot of opportunity to use both, kind of better ways of looking at the credit risk of an individual for a specific kind of credit, right. That gives you a much better sense of the risk as well as bringing in things that aren’t just your credit history that help me understand maybe more forward-looking, what your potential looks like and how much risk they’re is there. That’s not maybe portrayed in your history of credit.

Caleb Stevens: So, let’s start here, you’re applying artificial intelligence to the problem. Okay, we hear AI all the time, I remember 10 years ago I was sitting in a class in high school actually and I had a teacher who was really into AI for some reason. He was my Spanish teacher, why did he like AI, I don’t know, but he took a whole day and gave us this whole sort of dystopian vision of what the future’s going to be like. And it is sort of an Elon Musk kind of, you’re going have a chip in your brain and very sort of dystopian. When we hear AI, we may think of that, many of us may think is that little chatbot that appears on my app or I visit a website and it’s hi, I’m Dave, how can I help? Or I’m calling my cell phone company to get routed to the proper person to talk to, its most basic fundamental level, helps us sort of define AI and the way that you guys are, are using it.

Jeff Keltner: Yeah, I always kind of shy away from the term AI, although it’s hard to avoid these days because I do think people from the movies, they think of the matrix or they think of space Odyssey or 2001 and how the red light, and they think of some kind of human-like intelligence maybe in a humanoid form. And the chatbot is the same thing, it’s like a fake person sitting behind the keyboard, and we don’t use AI in that real way at all, that’s why I often use the term machine learning instead. It is really a way of saying; can I look at historical examples of something and think about how maybe I can make an estimate of something I haven’t seen before being true or not. This could be image recognition, I feed you a bunch of images of cats and dogs and I give you new pictures and go, can I use the old images? I told you, which ones were cats and which ones were dogs. Can I train the computer to figure out this is a cat, a dog?

This is actually a really simple way to say, hey, I’ve got previous history of something I want to make a decision about to what something is today or what might happen in the future. Can I train the computer to look at the old stuff and make predictions about the current or the future stuff? Anybody who’s used linear regression, that’s what this is, right? So, you take a linear regression, FICO scored a loss on the graph you draw a line that is machine learning. It just happens that this state of the art and capabilities of machine learning over the last 20 years have really exploded. And so, you can do much more sophisticated things now than you could then, but I really think of it as that very simple prediction problem, as opposed to the more general intelligence kind of, you can ask it any question it’ll respond. This is really just saying, can we take a specific set of questions and try and answer them more intelligently by looking at historical examples.

Caleb Stevens: This is not robots taking over the world and the whole……

Jeff Keltner: And if you have a simple scorecard, you’re making very hard rules as FICO’s good, maybe it is. The same thing like if you’re looking at that’s incomes or payment income ratios, like a high FICO doesn’t mean you can afford a particular payment on a given month, right? It may mean you have a history of paying bills really well but if you’re asking for a mortgage or a car payment, that’s outside of your free cash flow, like your history of paying bills, you need to be looking at both of those things, right? And if you do have a lot of free cash flow, but a history of defaulting on credit that increases the level of risk. So you have to be able to look at all of those things. You can do that in a human judgment sort of way, which is not probably the most accurate or fair or you can do it and what machine learning really does is say, which of these things is most indicative of in history, people who did or didn’t repay, why do you use a particular variable in the model? Well, because it turns out that it’s highly predictive of when somebody does or doesn’t repay and that’s the beauty of machine learning, it’s looking at history in this case loans and repayment behavior, and it’s just making the most accurate representation of what historically has told you, correlates to likelihood to repay, which is really what you want to understand.

Caleb Stevens: And long term, you’ve got more customers, you’ve got a better credit worthy base of customers and it’s an overall win I would think for the borrower and for the financial institution

Jeff Keltner: To a financial institution, it really means for any given risk tolerance if I can accurately predict how, what the risk of a given loan is, then I can maximize the number of people I can approve within any given risk tolerance. I can say I want to lower my losses; you can have an AI model that will say let’s lower losses, but I can probably still approve more people. Or maybe you say I’m really happy with where my losses are, but can I double the number of people I can approve and still keep those losses there. Can I take that if I have just thought of it today, I have all hard cut let’s say 660, 680, I don’t go below 680 on the credit score. You know there are good borrowers below six 80, that’s 660 to 680, that might be too risky for you at 5% losses but you know, 80, 90% of those people will payback. Can you find the 80 or 90 that are going payback and not lend to the 10 that don’t like, that’s the secret for machine learning. And so it means any financial institution can expand the customers they serve while controlling very precisely their risk and for consumers obviously, it means a bunch of people who didn’t previously get access to credit can and generally it also means we can lower rates because if we better understand risk, we don’t have to charge as much of a risk premium to make the interest income we need as a financial institution, so you can both make the system more inclusive and actually more safe and sound for the institutions.

Caleb Stevens: Do you see any carryover in terms of machine learning into the commercial side? A lot of banks obviously do consumer lending, but from any of them right now, they’re chasing CRE commercial loans. Has that sort of made any inroad into that world of lending and how so?

Jeff Keltner: I haven’t seen it yet, I think the bigger the deals the less these kinds of technologies are being applied in certain ways, because a lot of them are being applied to automated decision making, and those are things that are done in smaller size transactions. But I think it’s inevitable If you just say you can think about automated valuation modeling as a simple example that will come to mortgages and will come to commercial real estate. You can think about AI models now that as you start to do commercial lending that looks at the cash flows of a business, right. The way we’ve traditionally done that is either looking at PNLs or merchant cash advance is really a cash flow-based loan, right? Like we’re just taking year-old cash flow we’re basically fronting it to you a little bit. But if I can understand your seasonal cash flows, the business, the industry trends in your area, that’s a lot of data points that today we ask a human to kind of combine and over time, it’s just inevitable to me that machines will become smarter at being able to make the right inferences from that sets of data and make better risk predictions. And so I think, it’s just inevitable that all forms of lending are going to go this way, it’s a question in my mind of when not if AI and machine learning ends up becoming large portions of how we underwrite every kind of credit risk.

Caleb Stevens: Well, tell us about your own podcast because you’re in our world too, you host a weekly show each week. Tell us about your show leaders in lending and what all do you guys discuss on that show?

Jeff Keltner: Yeah, leaders and lending’s are our podcast kind of focusing on the consumer lending space. And so, industry leaders from banks and credit unions, a lot of community banks, and there some from fintech or people from credit bureaus kind of at the forefront of what’s the next thing in lending. There’s a lot of interesting conversation, I think digital transformation is top of mind for everybody who’s in the lending space, particularly after COVID. So, we have a lot of conversations about how that’s going, what works, what doesn’t, consumer trends for kinds of loans. You know, the buy now pays later space, how’s that going play with credit cards? These consumer personal loans are really interesting E-locks have declined, personal loans are up with a lot of discussions that are so kind of a weekly podcast with thought leaders in the consumer lending space about what the big trends and topics are.

Caleb Stevens: Well, with where you sit from a FinTech perspective, for the community bank executive who’s listening right now, who’s running a bank say between, I don’t know, 500 million to a billion in assets pretty small means a lot to their local community, but wants to stay independent long term, doesn’t have any plans to sell in the near future, which you know, more power to them because every day I feel like more community banks are getting acquired because there’s a lot of challenges that are facing community banks. We have a lot of shows and guests on here that talk about M and A and that whole world, any advice to those folks that really want to maintain a sense of identity, care for their communities, care for their customers, but they also realize we guys, we cannot get stagnant with our technology, with our digital transformation. We’ve got to be forward-thinking; we’re not going win the arms race with JP Morgan or Wells Fargo or any of the big banks, or even maybe the regional banks, but we do need to, the best we can stay up with where things are going technology-wise, any advice to those folks that you might have for them?

Jeff Keltner: Yeah, so I have two pieces of advice, I guess for those folks. I think it’s a really fascinating space, there are a lot of community makes that do want to stay independent. My two pieces of advice are number one, there’s a lot of FinTech partnerships out there that can help you is the gap and it’s interesting that we talk about FinTech, but frankly, your core systems provider is a FinTech. I mean, I may not think of them as a new FinTech, right. But the idea that you’re going find technology companies that build the best class of services and use those to provide quality programs to your members, that’s a real thing and it’s not new in this industry. I think people often think FinTech partnerships are new, I go, ‘man if you got a FIS or an FI Server or a Jack Henry core, you’ve been in a FinTech partnership for, for decades maybe.

But I think there are certainly new entrants who are providing more digital experiences, more up customized experiences, and a look at those because you’re right, you’re not going compete with JP Morgan’s investment in technology, but you don’t have to, right. You can leverage the investments being made by FinTech, by technology companies who are scaling their own investment in these kinds of experiences. I would really tightly look at that and the second thing I think people often have kind of a misconception about how to provide the best digital experience and so I would say the community banks have a real advantage in the human connection and the trust that they’ve built with consumers. It’s really fascinating because I talk to consumers and I talk to bankers, some people think the world’s going to the neo banks, no human touch, no branch, it’s all going away. I’m not a believer in that, I really think there is a place for human interaction and I think consumers, crave it, small businesses, crave it because they need help and advice. And so, what I think we get wrong is the sense that we have to digitize everything. I think what we need is to, to of deploying digital technologies so that the transactions, nobody wants to call somebody to ask the balance in their checking account. We just don’t right. I don’t really want to walk into a branch to fill out a piece of paperwork to apply for a loan. I want to sit down in front of my computer and be able to just do that in my, in my PGAs, on the sofa, on my phone. So, you need to have digital technologies to do that, but if you can also pair that with the ability to quickly and easily talk to a human cause when I go, hey, I need a mortgage.

I don’t need somebody to fill out the papers for me, but I want to understand, I hear about these arms. What’s an arm and a fixed, what’s a 15 and a 30, why should I pick a 15 or a 30 or an arm? Is that a good product for me? If you can be the person, they turn to at that moment that helps them understand the product mix. I would want to call somebody and go, hey, help me pick the right mortgage product. And once I’ve picked it, I don’t, I don’t need to spend time on the phone, filling out paperwork, like let me transact digitally right in my own comfort and convenience and, and timeframe, but make sure that when I need a through to a question or help to make a good choice, you’re there because consumers, business owners are craving that helps in navigating the products that are out there.

They don’t, should I get a HeLOCK or a personal loan for this I’m doing, I got to put a new HVAC in my house. Like how do I borrow money? Do I put on my credit card? Should I get a loan? Do I get a home improvement loan? How’s that different than a HeLOCK I need help there. And so, if you can really marry the trust, and that advisory role that you can play in a human way and frankly, leverage technology to get your people out of managing the process. You don’t need people managing paperwork, let the computer do that and let your people be value-added advisors and if you can do those things, then I think those community institutions have a really rich future. The ones that can find the balance of leveraging that human touch. That’s so unique that they can provide while also providing kind of a really easy digital experience when somebody is ready to transact and doesn’t, and I don’t want to wait on a phone call hold for somebody to be there, to take the app. Like that should be easy, if you can marry those two things, I think there’s a real opportunity for community-based institutions, to really thrive in the future.

Caleb Stevens: I think that’s great advice remove as much friction as you can through technology, but never sacrificed relationships, the personal touch, trust those things.

Jeff Keltner: Frequently enhance it. I mean, if you can stop having your people doing administrative tasks.

Caleb Stevens: Yeah, yeah.

Jeff Keltner: And picking up the phone, to do basic processing, then they’re more available to do the value-added advisory work. That’s so valuable. And that, that, that builds that trust. So, I don’t think these things are like an opposition. I think they’re actually working in concert when done and I think those community institutions that do it right will really have an advantage in the future.

Caleb Stevens: Well, I think that’s encouraging to kind of end on that note because I think a lot of folks can hear talks about FinTech and think man, human elements going away, this is not what it used to be. I missed the old days when I had a relationship with the folks at the bank. And what you’re saying is that’s maybe not the future at all, what we’re doing is actually removing barriers, removing friction at the same time, we’re still here to serve you and to have a relationship, but we also don’t want, you have to fill out piles of paperwork when you could easily, right-click a few buttons in PJs, on the couch.

Jeff Keltner: That’s right. I heard of one of the bankers on my shares. They said you need to make it easy to talk to somebody and easy to not talk to somebody. Yeah. Depending on what you want. And if you can do that, make it easy to, to not have to talk to anybody when I don’t want to. And also, really easy to talk to somebody when I do. I think that’s the magic for our community banks in the future.

Caleb Stevens: I don’t know if you know Lee weathering or not. He’s a chief strategy officer with Jack Henry. I think that’s his title, I might be wrong, but he said he was in a, he was with chatbot one time with some kind of app and the app told him he was 40th in line and he’s like, you probably don’t need to tell me I’m 40th in line. That’s not making me feel any better.

Jeff Keltner: Says just go away. So, you’re not going to get there. Yeah, if you can automate those boring tasks, if you can use technology to offload that, then you don’t need 40 people to be in line. You can make it immediately accessible to your team. And I think that’s what consumes customers want that at that moment.

Caleb Stevens: Well, if folks want to learn more about what you do and engage with the services you provide to help bankers across the country, how can they find you?

Jeff Keltner: Well, you can find me@jeffatupstart.com. So, I’m pretty easy to get ahold of, or you can subscribe to the podcast called leaders in lending anywhere you get your podcasts, and we’re just on the web@upstart.com. So, kind of come find us or drop me a line and we’re happy to have a conversation and discuss what we could do for you.

Caleb Stevens: Well, we appreciate your insight and have enjoyed the discussion. Thanks for joining us.