Starting a Bank in the Age of AI

Speaker A: Foreign. We're back in the studio. It's me, one of your hosts, Al Dominic, with my friend Steve Williams back

Speaker B: at HQ in Scottsdale. Good to see you, buddy.

Speaker A: Back in hq. And today we're going to do something I think a little different than previous episodes. You know, we've had some great bank CEOs share how they're trying to position themselves for future growth and success. And I thought let's take a pause and come up with some prescriptive ideas for building a bank on our own if we were given that opportunity. Right. There's a lot of things that are happening and so much of it is AI driven and tech focused that I thought if Steve and I took a few minutes to kind of get plugged in together around where and how we might try to build a really valuable franchise, it could be informative and helpful for our listeners.

Speaker B: Interesting. You know, I heard a stat, uh, I don't, I forget the name of the company but like for the first time with AI now we've had single person company create a $1 billion market cap and I think about all the great financial institutions that hunkered down for 75 years to create $1 billion market

Speaker A: cap and now one man, one man

Speaker B: with to rule them all. Yeah, and I know, uh, that's not the norm, but I think it's interesting to think about the shift in value of everything we're talking about. You know that song that goes Mad World? I think, oh, we're kind of in one right now. It's really, it's one of those moments every couple decades where there's a big kind of wake up and adjustment. I, I, you feel that this year,

Speaker A: what's funny is as you're singing, I, I took some liberties in advance of this because if we're talking AI, we got to talk about some of the tools that we're both playing around with. Uh, on LinkedIn, I've been trying to be a little bit more open with the exploration that I've, you know, kind of taking myself on with the uh, vibe coding front. So there's an app called Kiro which I know some fintechs are looking at in terms of their tech stack and trying to figure out if their code is really strong and solid. And I thought if I could start to dabble in this, that would be great. You and I have both used a little music app called uh, Suno and I just, I, this is for you. I thought we've got a great plugged in but we need maybe some new Music. So tell me what you think of this thing.

Speaker B: Okay,

Speaker A: so let's go.

Speaker B: We hanging around the studio. Oh, you've gone. You've gone.

Speaker A: I want to see some action.

Speaker B: All right, all right.

Speaker A: I could just go all day.

Speaker B: You got a Nashville. You got a new country version of Plugged In.

Speaker A: Well, you know, it had to have some fun. And it. I think that's the type of thing that shows how creative you can be in a very short period of time. And so the reason I thought this episode of Plugged in might be focused on if I was to build a bank from scratch today, what would I think about and, you know, what would I maybe walk away from? It's because folks like you and I can actually scratch that inner creative itch to say, hey, if I want to create a song, I now have the ability to do so. I don't need a band. I don't even need a studio. I can just get something that might be, you know, kind of fun. Same thing with businesses. And so, you know, we have so many clients that are trying to figure out how and where do they participate with some of the newer technologies that are coming to market.

Speaker B: Right.

Speaker A: As they think about the outcomes that they're driving towards the delivery channel that they already have in place and how they're trying to expand them. So I just want to reverse engineer it because we bring no, you know, tech debt, no legacy mindset to this conversation. Um, um, we're not really scripted. We've both been traveling a bunch, so we've got anecdotes I think we can throw into the conversation. But, like, if I was to start a. A bank today, I'd basically look at you and say, what are we really good at? And, you know, what is it that we can do that will move the needle for our potential customers?

Speaker B: Right. And, you know, I think a headline recently, uh, you've talked in the past and written about nubank out of Brazil, which is also in Colombia, Mexico, now coming to the US they had a milestone of exceeding 100 million customers, which is, I think, the first digital bank to hit that number. But if you look at where their CEO is talking strategically, it's. We nailed it as a digital native bank, and now we want to be an AI native bank. And they were ahead of the curve in acquiring a company called hyperplane in 24 two years ago to say, we want to kind of now shift from digital to AI. Um, as a lot of banks, you know, rightly so, are still kind of trying to get up to speed on digital maturity. So I think it's a good time if you look at investors, whether it was the price of Salesforce down 40% in the last couple of months or private equity assessing every portco right now to say what is the disruption to enterprise value? We need the same kind of stress testing, pressure testing for banking right now.

Speaker A: 100, 100%. Because think of it like this. If there was a well funded AI native company that came in and said I'm going to target your top 20%, you know, customer base like the most lucrative, most profitable, supposedly most loyal with something that's a compelling offer, what's going to keep them from not at least taking a peek and if they're interested, making a move.

Speaker B: Yeah.

Speaker A: And so you know again I think about some of the different things that are out there. Some of the AI tools that I've been just dabbling with just in case you're interested. You know, I said Kiro, we just listen to Sono uh for music but there's you know, co work we have Zapier Whisperer. From a visual standpoint you have like Nano Banana and things where funny words right. To have to say out loud. But these are some really valuable learning opportunities for us all to look at and say if I had to start something from scratch, what would I do? And I'd be using tools just like this.

Speaker B: Yeah.

Speaker A: Um, because again if I can get smarter, faster on where there might be some very discreet opportunities to win in the niches. You know we've been writing about how you find the riches in the niches and AI is going to help us.

Speaker B: Right. And for the Bezos, uh, your, your margin is my mission. So if my, if, if a top tier commercial bank had some disruptor, I think one thing I, I would start with that the startup advantage would be I would start with a data model. Like let me build a foundation of data to think about. Okay, who are those commercial businesses? What, what how would I manage relationships, credit? I would build out the data model. So all the things I wanted to build on top of it were nice and clean. And as we know uh, for a lot of banks right now just cleaning up data to be the foundation for AI is a big one. So I'd start really thinking that through.

Speaker A: I'm just laughing because I can tell you've been a consultant for 20 some odd years because every consultant starts with the data. But you're right, you know, having a strong framework as to who and how you want to serve, how you're going to differentiate what you're bringing to them. How can you provide that experience that they don't necessarily know that they need, but the tech lets you identify some patterns pretty fast. And so again, it's the ability to move, which is important, but to be able to find clearly using the data, where there might be a niche, uh, market to participate in. That's kind of how I would think about starting a bank. I wouldn't worry about the charter. I mean I can buy a charter. I would rather think, do I get a small tiger team of very curious individuals who can understand the business applications that can be pulled together to create something of value. I'd uh, bring some folks who understand the regulatory and compliance nature of the banking space. But I could have a very tight team that says if we get some real strong data to begin, the tools are going to help identify some opportunities that maybe we wouldn't historically look at.

Speaker B: Well, think about. Also I'd be thinking about, hey, we collect a lot of information to be your lender and we also are, uh, where we facilitate your payments and deposits. So between all that information, I could build a lot of intelligence to every day let's monitor my commercial customers for what I need to do, credit risk and payments. But I might be able to spit things back to them that are valuable through digital. So you know, if I could, if I could be personalizing and providing agents to do things for my commercial customers through my treasury platform or it would be a great way to think about how do I add value more as that trusted adviser.

Speaker A: Well, and you bring up agents. I was in Nashville the last few days, which is where that music, you know, kind of came from. Um, I was talking to a handful of bank executives that run institutions between, uh, I'll say like 2 and 5 billion in size. And they are all curious about agenic commerce and what's going on, but they don't know really where to start. They don't know how to create their own agents. They don't know who's creating AI agents to help them. And so again, if I'm thinking about you and I starting a bank together, I'd be building our, our brand and our expectation around speed and using, you know, the ability to take some of these agents, have them built into the way we operate from day one. So instead of trying to re architect existing processes, I'd be building processes on top of the, ah, yeah, you know, investments that we're making.

Speaker B: I think it goes with what we're seeing with, you know, everybody talking about, well, the Trump administration has been very friendly and with the genius act and stablecoin, you see this mad rush to everybody wants a bank charter.

Speaker A: Yeah.

Speaker B: Um, so either they're applying or acquiring or looking at trust powers. But I think the theme there is the techies are absorbing banking knowledge faster than the bankers are absorbing technology.

Speaker A: Sure.

Speaker B: Uh, and that means I would, if I was starting from scratch, I would have a tiger team of techies who aren't constrained by. I grew up in banking, but, you know, let me give you the problems to solve and let's go from there. And then, like you said, I'll bring in my KYC compliance person. My, my. Someone with wisdom on credit. But I think the challenge would be I would start with a team who could just, you know, think fresh about tech.

Speaker A: Yeah. Because, uh, I mean, this is where you don't ignore history's lessons like it's people product's performance. And if you get the right team of folks together with a, ah, shared vision. I mean, I think back to our conversation with Mike Daniels at Nicolet, when he talked about the culture that had been built out in Wisconsin as one that really reflected a team that could lock arms and know exactly how and where they wanted to serve their market to the best of their ability. And they want to win at the upper end of the folks that they serve, and they want to do that consistently. So I'd be looking to recruit folks who have that same desire to win, who can lock arms and not look at each other in a competitive fashion. But to say we all bring different skills into this story, but we are all aware that some of these AI, um, tools that enable us to perform differently have to be fundamental to how we do things.

Speaker B: I think all the things around security and of course, you know, cloud being wonderfully popular right now with, with their, the platform. But then also Claude taking down a company this week, I think that what I would do right now as a banker is make sure my, my CISO and my techies could provide a safe space that's not connected to the whole enterprise for some of my smart people to start playing with agents. We have something at Cornerstone, you know, called AI Day. We've done it for four years where we just say, how are we doing with AI? Let's showcase some things. And we had three of our team members who built their own agents, kind of show off what they were doing. I like to say that the team member who won the competition was the oldest in the crowd. But you got to find a safe space to let some of that Experimentation happen right now, like very hands on at the individual level. That's one thing I'd encourage those folks who say, you know, how do we build the agents? Well, you probably have folks who are playing with stuff at home or they want to play with it at work, but they, they kind of need the ground rules and they need a sandbox that's safe while you do all the infrastructure, data, security, and take that stuff very seriously.

Speaker A: Well, and this is where again, if I was starting a bank, I'd be recruiting really smart young people who are just living this every single day in a way that's different from you or me. Right. Who's had, you know, 25, you know, 30 years of business experience. We can help teach the business of banking.

Speaker B: Right.

Speaker A: Where we can be taught is how some of these, you know, digital tools can be manipulated in a, in a clever and creative way. So, you know, again, I think it's important that we think small teams that move with a sense of speed and purpose. Um, is not to break things randomly. It's to give everyone the opportunity to say, we've got to all experiment every day. And when you think about experiments, I, I was struck with Jay Sudhu's experiment with OpenAI the other day for his earnings call. I don't know if you saw this actually you. Did you text me about it.

Speaker B: Yeah. It was actually Sam though. Was it?

Speaker A: I would say. I'm sorry. It was him. Sorry, Jay, I didn't mean to toss you under the bus.

Speaker B: Yeah, he was, uh, kind of, uh, letting his agent do the tr. The earnings call.

Speaker A: Yeah.

Speaker B: And I think it's a way to say this is moving, it's moving fast. And even someone at my stature could do something like that.

Speaker A: And that's like a 26 billion dollar bank out of Pennsylvania. And for listeners who may not have heard this, Sam had earnings call for the first 20 or so minutes where it was his agent or his, you know, clone that was giving the results. And he used that as a way to introduce an open AI business relationship that they were proud of. Um, what's interesting about him, he's married to a woman who has, and I think she still does, works for the NBA. He and I had a chance to get to know each other a little bit a few years ago and he was telling me about the NBA having these big splash moments and how he always wished that a bank would have a big splash moment. So I'm tipping my cap because that was a pretty cool thing. Now for a new bank, I'm not sure that I would want an agent, uh, to represent us as a leadership team. But I think using some of that creativity certainly comes into play. Um, if we're talking though about performance and outcomes, you know, we have to think about what are we going to measure and you know, are there things that we think create value over time that some of these new, newer tools could help us uncover?

Speaker B: Well, definitely, you know, people are looking at the efficiency ratio, the famous, you know, what's going to happen there. I like to joke that, you know, average bank runs 250 to 300,000 in revenue per employee and an innovator like Chime is at a million and a half. So there's certainly huge opportunity to think about, you know, what is the revenue per employee that I can build in a bank as I put more IP and automation and agency work into the whole thing. Now that's interesting because, you know, I think what you would also do though as a commercial bank is think about the concept buzzword I've heard lately, which is service as a software, which means I am going to have that last mile of human relationship, but everything behind it is going to be software driven. Uh, we love in banking to have pride in our relationship. Managers, our credit officers, great. But everything behind them should be coming intelligently automated and so your expense structure becomes that human layer. Uh, going forward, of course you got some really sharp people managing the robots and designing the next agent and deploying and all that agile teams with human relationships in front is the future.

Speaker A: Yeah, and you know, this ties into the broader, you know, themes that we've discussed on previous episodes about, you know, the movement of money and where there's upside to stablecoin, especially on cross border transactions. You know, so depending on the type of bank we were to stand up, it would have to figure out, you know, where and how do we want to participate with, you know, tokenized deposits. What are we thinking about when it comes to stablecoin? I don't see us issuing our own because we would be a small, you know, startup that's aspiring to grow fast. But I think there's a lot of interest and appetite from community banks to better understand can they participate with some of these conversations? I don't know that they can unless they've really been clear about what they want to do and what they don't want to do. And the ones that have said, hey, we know where we're not going to waste our time, but we do know we might have some cross border opportunities to dig in there seems to make a lot of sense.

Speaker B: I've, uh, been in meetings recently where one of the things about all this kind of the collision of AI and blockchain stablecoin is banks live on a lower cost of funds than the brutal world of the daily market of money. Mark. So, and it's always been like, with information, no one's going to have money laying around. But, but they do. I mean, today we have banks with, you know, 50 basis point cost of funds right now. And it didn't happen when they deregulated national banking. It didn't happen with the Internet. You know, I remember Forester 25 years ago said, all dumb money will become smart money. Well, here we are, 25 later and there's still money sitting around. Now why relationship. I, I'm leaving it to generate things. So I guess where I'm going is the cost of funds is going to be even more embattled now with what can happen with stablecoin and AI and the intelligence it brings. However, the opportunity with that bank is add more value to the movement of money through smart contracts. Add more value with niches, uh, with the intelligence you can give back to your customers so they see the value and they'll, they'll let that money stay in the system, uh, going forward. But I think more than ever, this one scares people about cost of funds. More than mobile, more than digital, more than interstate branching.

Speaker A: Yeah.

Speaker B: Than, uh, we've seen in the last 50 years.

Speaker A: Well, you know, uh, we'll pivot kind of towards the end to say, hey, we could build a bank. But I'm not sure that's where we really should be using our time or talent. But I think from an intellectual standpoint, it's helpful to say, if we had no guardrails, what could we do? Obviously we've got regulatory expectations those are shifting. We have compliance, uh, responsibilities where we have to make sure we're not doing something, you know, foolish. Privacy data, you know, concerns, we understand those, but the tech is moving at such a rapid clip that we would be foolish not to try to figure out, are there places that we can do better?

Speaker B: Yeah.

Speaker A: And so I, um, I jotted some notes down. I brought them because there's quite a few things that if I'm an established bank, like, what should I be thinking about? Because they don't have the luxury of you and me from a dead stop saying, hey, we're going to start from here. So if I was a, like a 10, $20 billion bank and I'm, um, not starting over, but I want to choose what to protect, maybe what to fix and, uh, also like, what to let go of. I'd start with, you know, defending my core business before moving out and looking at the edges and figuring out where am I slow or expensive or what could be done. You and I have talked and you're the one who really encouraged me to think about a kill switch for some of the investments that have been made in the past that may not be returning value.

Speaker B: Right.

Speaker A: So I'm curious, you know, that return on tech concept seems to bubble up again when it, when we think about AI's impacts.

Speaker B: I remember on our, uh, what's going on? Ron Shevlin asked me, is, can tech still be a competitive advantage? And I said no, it's alignment of tech with a very distinct business model, niche and strategy. The ones that really take their resource allocation, both capital and opex, and put it towards something that they, that passes the Michael Porter at Harvard test. We do things that are different than our competitors, that are highly valued by our customers. So to think about that and then build that with the new tools. The other thing I'd say real quick, Al, is that I think about my leadership team of the future. Uh, it's not just about I have wisdom and I can oversee things. I think the bank of the future will have people who design and deploy systems, models their design skills, and they have to know what the tech can do and they can work with techies, but they're not just managers of people in hierarchy. They're designers and deployers of systems that do things. I think that's the big difference you see in say, a fintech workforce versus a traditional banking workforce.

Speaker A: Yeah, I think those are the types of people that would look at data as an asset, not as, you know, just, uh, you know, something that needs to be cleaned up.

Speaker B: Yeah.

Speaker A: And so if you can start to, you know, use some pattern recognition more thoughtfully. And this again, was where some of these tools can help you see something that maybe you would have overlooked. You know, you can start to, you know, shift and allocate your capital in a more appropriate fashion.

Speaker B: Yeah.

Speaker A: Um, anecdotally I was with Kim Snyder, um, who started Clarice.

Speaker B: Yeah, great company.

Speaker A: And she was talking about the early days of that company. So there were evidently six banks that basically trusted her to start the company blindly. Um, and she was able to identify some very discreet challenges that a CFO say, hey, I need, I need some

Speaker B: help kind of data I crave and goes.

Speaker A: And so that's the mindset I Think that banks could continue to look at, which is not to start a new business like Kim, but are there discrete issues that are not being solved and do you have data that could potentially bridge the gap? And then how do you again treat that as a very positive piece of your business? Because we know banks are sitting on a treasure trove of information. It's just a matter of unlocking it. And so, uh, again, when I think about how I would encourage some of these banks that are a little bit bigger to re consider what's possible, it's certainly to go deeper with their data. Don't ignore the regulatory advantages that still exist. Be mindful of some of the legislation that's been passed that's also encouraging folks to come in that may not be,

Speaker B: as it's the Oklahoma run night right now, to come into banking. Absolutely. Based on uh, what's about to happen with AI and blockchain.

Speaker A: But that's where I think that banks sometimes limit their recruitment efforts in a way that they shouldn't. Because if I was a smart, talented person who's looking at industries that are going through really interesting periods of change, I would say I could go work for a, uh, you know, a chime, I could go work for a stripe. Or I could go work for a bank that is trying to reposition itself and actually has brand equity, customer loyalty, the ability to invest and allocate, you know, time and resources in a, in a way that makes my career that much more rewarding.

Speaker B: Yeah.

Speaker A: So I just get excited about some of this stuff.

Speaker B: Yeah. And I think you pair, I said this before, but pair up the, these experienced bankers who have knowledge and credit compliance, etc. Loan servicing with, you know, some tiger team folks who can help pressurize. Well, this is where this is going to go. You're, you know, how I take waste out or how do I just change it completely. I think two things to really think about too, about uh, AI is we have a lot of unstructured data in banking, but we can grab it now fast and get better insights with it. Think about the insights of every credit memo sitting out there unstructured to just have it go read and give me some stuff on, on trends recently. And you know, I think that's fascinating also just conversations like how people now can take conversations, whether it's in the call center or senior loan committee and turn it into continual insights. These are big like pragmatic ways banks could start changing how they operate this week.

Speaker A: Yeah. And I think on the individual side we can talk it but we have to walk it. And so there comes a point where you just have to allocate time so that you can get uncomfortable with some of these things that are going to impact your business. And the reason I mentioned vibe coding earlier is I've been in conversations where people get really excited, and I don't have as deep a, uh, you know, knowledge base as I would like. So how am I going to close that gap? I have to start trying it out myself.

Speaker B: Touches the clay, right?

Speaker A: It's going to be ugly. It's not going to be embarrassing.

Speaker B: Yeah, Touch the clay.

Speaker A: But if you're curious. And again, we talk about curiosity being a key attribute from a leadership standpoint, this is where you start to model it. Because if I saw you, Steve, doing something like we did with the. The music, if I can, you know, create something of value that a, uh, younger person says, well, if Al can do that, I can do that faster. Great. Let's learn from each other. And so I think this is a wonderful moment in time for bank executives to say what we've done in the past, you know, it needs to be protected, but we also need to grow and change in some ways that we are not really comfortable with.

Speaker B: Right.

Speaker A: But instead of shying away from it, let's just press forward and show what can be done.

Speaker B: Yeah. It's not a. It's not a big company. Partnerships, command and control. It's. Everybody get their fingers dirty right with it right now, jump into it. And you only learn by doing this stuff. But then you realize, wow, this is not the old world where you kind of plan and then design and build and deliver. This is like, plan, design, build, deliver, plan design. You know, it's happening at that kind of pace.

Speaker A: Yeah. No, and that's why I just thought, you know, if we're gonna hop in the studio together, we should talk, you know, in kind, uh, of concept terms and in practical ones. We've got some great entrepreneurs that are going to be joining us in upcoming episodes. I love hearing how they've built things, but how they're not satisfied with the status quo. And that's why we've got some past, you know, guests that, you know, like a Mike Daniels or Jay Hillenbrand who was just on from stockyards, who say, hey, we've got a legacy that we build on, but we don't just, you know, become the sleepy old institution that loses its luster. So just to continue to be creative is why we get plugged in as we do.

Speaker B: Thank you so much, Bud. To see you.

Speaker A: If you enjoyed today's plugged in episode, make sure to hit that follow button on, uh, Spotify, Apple Podcasts, YouTube, or wherever you're listening. We've got more great conversations coming up and you won't want to miss them. Thanks for tuning in.