The Path to Banking's Dual Workforce, with nCino CEO Sean Desmond

[Music Playing]

Jim Marous (00:11):

AI is shaping the way banking gets done. Most institutions still underestimate what AI means to staffing, decision-making, and customer engagement. New research from nCino shows that nearly 9 in 10 banking executives believe their institutions will operate in a dual workforce environment where humans and AI agents work side by side within five years.

Jim Marous (00:34):

I tend to think it's going to be a lot quicker than that myself, but only one in five can tie AI investment to revenue growth. Banks have spent the last several years experimenting with AI. The next phase is going to definitely be harder.

Jim Marous (00:48):

Today, I'm joined by Sean Desmond, the Chief Executive Officer and President of nCino at the nSight event in Charlotte, where it's a client event and sharing a lot about what's going on in the industry.

Jim Marous (01:00):

So, Sean, I just saw you on stage a few minutes ago. I'm going to start right from where you started, which is I think the hardest thing for the banking industry to do is to embrace the change that's upon us, and to actually immerse ourselves in AI because unlike almost any technology we've had in banking before this, this is a consumer facing technology as well.

Jim Marous (01:23):

So, you can get up every day and I know I do. I can go to my AI tool and ask it some questions or ask it for some solutions or say, "Hey, I have this going on, what's going on?" You talked about the fact that you as a leader at nCino said, "I got to get in this game. Even though as a product person, even with my background, I can't just talk the game. I could actually walk the game."

Jim Marous (01:45):

What have you done on a personal basis to make that happen? Because I think most banking leaders can really learn a lot from the dynamics of just getting into it full-bore.

Sean Desmond (01:57):

Happy to share. Thanks for having me, by the way. It's great to be with you. I do think so much about this inflexion point that we're at is attitude, for sure. And so, if you zoom out for a second, even from banking and from my personal experience, change is hard. And when people go through transforming the way they've done things for many, many years, they always need to have this sort of moment where it hits them personally because we can read about it in the news and in the media, which we do.

Jim Marous (02:34):

That's part of our AI injuries.

Sean Desmond (02:36):

We're getting hammered with it every day. And until it becomes personal, it's not real. So, you referenced the story that I shared on stage, which is really a moment where we've got an AI strategy here at nCino where we have embedded intelligence within the workflow built on a data foundation and we've been serving up AI solutions now for over a year to our customers.

Sean Desmond (03:00):

And I've been pushing our teams hard to deliver on the promise. And at the same time, I really challenged myself recently and said, "Well, what are we doing differently to run our company?" I'm telling everybody to help banks run their institutions differently.

Sean Desmond (03:16):

And so, I thought about my routine. I just thought about how I manage the business, the set of metrics, dashboards, KPIs, how I get that information. And in that moment, I just said, "Let's start now. Let's start now." We leverage Claude internally from Anthropic. Every employee at the company has a license for Claude, including the CEO.

Jim Marous (03:37):

That's a big commitment because that opens the door to some risks. But it's interesting that you say that. We'll talk about that a little bit later.

Sean Desmond (03:45):

It does. And you have to manage the cost and you have to get the outcomes, but including the CEO. And so, I said, "Alright." I sat down and I started iterating with AI the first draft of what I call our CEO agent stack. And so, instead of assigning that out, and actually, when I first interacted, guess what the recommendation was from Claude: have your Chief of Staff do this.

Jim Marous (04:10):

Interesting.

Sean Desmond (04:11):

And I said no. First of all, I don't have a Chief of Staff, I run the business. Second of all, let's go. Why not me? And so, we started iterating on this. It's essentially a spec. The same type of a spec you would take to a product team to go then go through an SDLC and take to production.

Sean Desmond (04:29):

And I was able to actually build that brief in under 90 minutes at night after dinner. And then I was able to immediately send that brief to two of my best AI engineers and said — I actually asked them, "Am I oversimplifying this?" And they're like, "This is on par with any spec that we would've gotten that would've taken three weeks to write by a requirement-"

Jim Marous (04:54):

And they would've had to send it up. As opposed to you, which makes it so that it actually is coming from your mindset and your experience, and then having them apply what they need to do.

Sean Desmond (05:04):

Absolutely. Just directly connecting with the business problem we're trying to solve. And the business problem we're trying to solve is how can I proactively and not reactively run the business? Some of the things that I want to know as a CEO every day are what every other CEO wants to know.

Sean Desmond (05:18):

First of all, I want to understand the competitive landscape. I want to understand who's coming for me. Actually, you heard Frank Sorentino on the stage challenge the audience there: "I'm competing every day and I'm going to win."

Sean Desmond (05:29):

So, what are my competitors doing? How are they thinking? I want to understand my customer satisfaction. I want to get signals from my customer base that if my net promoter score is down for whatever reason or a particular customer's survey ratings are down, or actually their usage and login and interaction and engagement with our platform is down, why?

Sean Desmond (05:50):

I want to understand my pipeline. I want to know if there's a deal greater than $250,000 that just moved out by 30 days in terms, it's called the close date because that affects my quarter. We're a publicly traded company, I need to do this well.

Sean Desmond (06:03):

So, all these things traditionally have been things that show up in a dashboard in hindsight. And now-

Jim Marous (06:10):

And a dashboard that doesn't give you answers to your questions about those dashboards. Now, you're feeding that, you're able to say," Okay, here's what I see. My customer satisfaction rating went down two tenths of a point. What do you attribute it to?"

Sean Desmond (06:26):

And it'll do the digging around your data and outside data and say," Well, it may be a competitor's just made an announcement about something that makes it look like yours is behind." Or, "Oh, by the way, you have some implementation that haven't gone as smoothly as you would've liked." But it's great because you have that immediate ask-answer, and you can test it, but the reality is as your data set gets better and stronger, you're able to ask questions, and it is addictive. We talked about that, isn't it?

Sean Desmond (06:55):

You have to stay focused, and you have to understand what outcome you're looking for. I think two things that are really fascinating about this new way of running the business. One is before I have my coffee in the morning, all the information is right there for me, rather than having to refresh and click and refresh the dashboard.

Sean Desmond (07:15):

Number two is I can interact with it. So, in that scenario where I talked about, let's say I'm looking for velocity in my engineering. And so, my CTO tells me we've actually increased efficiency in the engineering team by 60%, and that's amazing, and I want to know why. So, I can actually drill right in and I can interact with the agent.

Sean Desmond (07:36):

And by the way, this is an agent stack that is informing every department in my organisation. So, you have agents, you have sub agents. The sub agents for product, sales, service, legal, finance, HR. And every single one of my leaders in these departments, they can do the same thing. And so, we're doing it in unison. It's pulling the team together in a direction where we're kind of-

Jim Marous (07:56):

Without having to have meetings, without having to have the virtual calls.

Sean Desmond (07:59):

Thank God.

Jim Marous (08:00):

I mean, think of the time to save, plus everybody's working from the same data set. So, a couple of things I sense from that that many financial institutions that you serve aren't doing quite as well. Number one, they're not applying the access to AI across the board (many, most aren’t). Secondly, the sharing of data among your teams and among your stacks to make it so that what you're seeing on your dashboard is also what they can see on their dashboard.

Jim Marous (08:29):

I'm sure in your evolving knowledge of how to use AI, you become not as much the answer man as the question man. And that's a transformation just as far as leadership where you're usually being asked for the answers. Now, you know that your power is really in the questions you ask and the answers that can evolve from that. But again, to be able to serve your clients the best, you actually have to be doing it in a way that you're recommending your clients do it.

Sean Desmond (08:58):

You have to walk the walk.

Jim Marous (08:59):

Your research that you just came out with, which is talking about the correlation of AI and humans working together, which by the way, a link to the research is down below in the episode notes, but your research made it very clear that 9 out of 10 institutions are working with AI. Much of that is still in the experimental stages as opposed to transformation.

Jim Marous (09:23):

That hurdle is no small hurdle, and unlike other transformations we've had, this is not coming at a pace that we can actually deal with. Even mobile, as fast as it was, hasn't come at us as fast as AI is, and the consumer now has a lot of options in the marketplace, not only within the financial services industry, but outside the financial service industry to actually see how it should work. So, I talk about it as being like the GPS of financial engagement with the financial institutions and their clients.

Jim Marous (09:55):

How are you working with your clients to bring them up to speed knowing what they have to do? Because many of them aren't going to be able to hire AI personnel that are going to be managed in the AI process. They're looking for their partners to bring that skillset to the marketplace. How are you bringing it so that you can be that AI tool next to that human?

Sean Desmond (10:14):

And we absolutely want our customers to hold us accountable for being that AI partner. And the pace and the velocity is like nothing we've seen before. It's fun, it's chaotic, it creates anxiety. It creates some fear, uncertainty and doubt in the marketplace, and it really does really amplify the gap that we see today between investment and outcome.

Sean Desmond (10:37):

And so, what you see is, to your point, everybody experimenting with AI and it ain't free. So, you have people spending real money on AI. You don't see them correlating all the way to the outcomes because a lot of them are either in a pilot phase or a POC phase. And in fact, they might be doing this with only test data because their banks still have some sort of a governance policy that says you can't do it with production data.

Sean Desmond (11:02):

So, it's all by design in a sandbox until they're ready to flip that switch or take a signal from the OCC, or take a signal from somebody that says it's okay. And here's what's really interesting about it.

Jim Marous (11:15):

Is that an excuse or is that real?

Sean Desmond (11:18):

Somewhere in the middle. It’s somewhere in the middle.

Jim Marous (11:20):

That's what gets me, is that we look for the excuses that we've used in the past, and the reality is we have to push the regulators because the regulators are as far behind as we are as an industry, and that is … I'm sure as a server and a supplier to the financial service industry, you got to push them into that discomfort zone a little bit.

Sean Desmond (11:39):

You absolutely do. And listen, regulation matters, compliance matters. It's all there for a reason. Our industry has learned lessons over time the hard way, and we're trying to avoid that, and I respect that. At the same time, you will have two different banks at this conference that are regulated by the exact same policies with different postures toward AI.

Sean Desmond (12:01):

And so, it is somewhat nuanced and in the middle, and it's cultural and it's some attitude, but you have to be compliant. And that's where I'm really, really bullish on the position of nCino because for 15 years, customers have trusted us with their data. We were built for the regulator. When you look at the first generation of our workflow embedded in the things that we do across onboarding, account opening, loan origination, as well as credit monitoring, those are all built with regulation in mind.

Sean Desmond (12:33):

With compliance in mind, we've never compromised that position of trust. We've never had a security breach, and we have a CISO that's in lockstep with our Chief Product Officer that understands our customer's expectations.

Sean Desmond (12:46):

And so, what you do see, and we're in the envious position in this market that we have customers that have trillions of assets under management, the largest enterprise banks in the world, and then we have community banks, IMBs and credit unions on the corners of Middle America that maybe have a billion to $2 billion.

Jim Marous (13:03):

They have the same challenges though.

Sean Desmond (13:04):

They have the same challenges, and they have the same regulation. And you see that risk profile change as you go up and down market, and you also see the skillset and the attitude and the posture toward, do I want a vendor simply to serve this experience up for me, unpack it and deploy it hands off, or large enterprise bank might actually have an army of engineers in-house.

Sean Desmond (13:29):

And what they want to do, is they want to experiment with agents they're building side by side with yours, and have those interact because there are things that we do in our lane, and there are things that they're also doing maybe in a call center experience that we don’t focus on.

Sean Desmond (13:42):

And those agents and sub agents need to talk to one another. And so, we're building the platform as an agentic operating system that you can do both of those things up, down market with compliance, and yes, we're going to push our customers and we're going to challenge them and remind them that we haven't compromised their trust in 15 years.

Jim Marous (14:01):

Well, it's interesting because we tend to look as bankers for the answers that serve our own narrative. So, there's click bait out there that says AI is not bringing the returns that were expected, when really, if you read further, you realize it's just a lack of measurement as opposed to it's working that we're not paying off. We are and we're doing it in little piece because we're still testing in many cases.

Jim Marous (14:26):

And the challenge then is how do we get people beyond the experimentation stage and even more so, how do we get them beyond back-office efficiencies only? Because what we're seeing in our research is we're seeing a lot in the fraud and risk area, which makes all the sense because that's where AI really started.

Jim Marous (14:46):

When we start looking at agentic AI in the dual workforces that your research showed, we're really finding that a lot of it has to do with offloading jobs from employees that were repetitive. Not necessarily getting rid of the employee, but raising up what they can do.

Jim Marous (15:03):

How do we move away from only what I'm going to say the lowest hanging fruit, the easy wins, which are, okay, let's cut workforce or let's cut jobs, or let's make it more efficient so that our costs are going down. How do you, as a supplier to the industry, not only support that, but how do you push them into that discomfort zone of the customer experience, where the real opportunity on the revenue side can really pay off?

Sean Desmond (15:29):

There's really three things that come to mind. I mean, the first thing is culture and attitude. You simply have to challenge people to get outside their comfort zone and jump in, and everybody has to be in. I think the folks that are most at risk are the folks that are opting out of actually leveraging AI when it's available to them.

Sean Desmond (15:47):

And then the second thing is you do have to demystify the whole compliance, security risk. We have the same regulation, but banks all have different internal policies. I talked to an AI leader recently who said, "I'm going to share with you our policy and our governance. It's fully documented. And if you can build the APIs and connect us, you're good." And actually, that breaks down a wall because while their policy, maybe it's more rigorous than another, it doesn't really matter, I can build to it.

Sean Desmond (16:17):

I talked to another bank, same size, same scale, actually same geography. And so, I asked the question, can you share with ... they haven't documented it yet. So, they don't even know what compliant looks like for a vendor like nCino because they haven’t-

Jim Marous (16:31):

Or for AI in general.

Sean Desmond (16:32):

Defined it. So, we need to define that together. And then third, I do think that from a dual workforce perspective, you have to be able to aggregate what I call the micro wins up to the end-to-end results. Had a customer on stage this morning and we're talking about 50 to 60% reduction in credit review cycle time. We're talking about actually cutting existing hours by 1000 hours, a half, like the 20/80 rule.

Sean Desmond (17:06):

So, if he can save a thousand hours on his employees doing the same thing over and over again, and he can repurpose those folks toward where? The customer. And as much as everybody likes to talk about relationship banking, how many jobs in an institution are far from the relationship?

Sean Desmond (17:25):

I think this brings us all closer and it's exactly what I say at nCino to everybody on my team. I don't care what level you're at in the organisation, get closer to where the work gets done, AI is democratizing all that. It's fantastic.

Jim Marous (17:38):

And your job is to become a hero to your clients. So, if you can help them know their competitive environment better than they would've been able to find themselves, if you help them understand their customer base better than they're right now ready to do, if you can bring them that level of comfort and do it in a somewhat of a hands off basis for them, it works to their advantage. It gives them a little bit of both worlds.

Sean Desmond (18:01):

Your organisation structure is changing based on AI as well. I mean, the implementation of tools and the human plus AI, how does that change what you're asking areas to do and how to implement from an organizational standpoint?

Sean Desmond (18:16):

So, I think it first starts with the KPIs and the metrics that you use to measure success. And I think you can be a lot bolder on the expectations with those. We live in a world today where if I talk about product roadmap, anything outside of three months is a lifetime. We used to talk about product-

Jim Marous (18:37):

Okay, so you came from product. That used to be a year, two years, three years in some cases.

Sean Desmond (18:42):

Exactly. 18 months, people would serve up the roadmap in 12, 18, 24 months. I would be embarrassed to show that right now. And by the way-

Jim Marous (18:53):

You’d be out of business doing that.

Sean Desmond (18:53):

24 months from now, the thing that we thought we were going to build may not be relevant anymore. So, I think you have to understand that you're going to set different expectations with velocity and scope in the implementation deployment world, because we're re-imagining everything from how we can integrate data to how we can actually deploy and unpack our agents.

Sean Desmond (19:13):

Guys, this stuff is rolling out in weeks to months versus months to years. At an enterprise bank, it was not uncommon for folks to have multi-year journeys to do digital transformation for complex commercial lending. We're doing that in inside four months now.

Jim Marous (19:29):

Crazy. Let's take a short break here and recognize the sponsor of this podcast.

[Musci Playing]

Jim Marous (19:37):

So, when you're working with financial institutions today, what is their biggest holdback? You're trying to push them into areas of discomfort, but you're working with the top of the organisation with the people that are most tech savvy. They may like what you're selling, but they're not implementing.

Jim Marous (19:57):

What gets in the way of actually implementing? Because I'm sure most of the people that are listening to the podcast have the same challenges. They go, "It all sounds great." Yes, all apple pie and all the good stuff, but the reality is it's not always that easy. What's the biggest challenge that you see?

Sean Desmond (20:13):

Listen, the more things change, the more they stay the same so to speak. I think the biggest challenge is actually change is hard re-imagining work. Yes, risk, security, compliance, those can all be excuses, but we can overcome most. What you can’t overcome is a culture and an attitude of an institution that is resistant to change.

Sean Desmond (20:36):

What you can't overcome is a leader in the top of the house that says, "This is all great, but I'm a few years from retirement, so I'll let the next guy figure it out." What you can't overcome is that top-down directive is not the AI way. It's much more of a bottom up, and so you have to understand that you have to empower your people and let them know that they're empowered to experiment.

Sean Desmond (21:01):

And yes, you have to control the cost side. I don't want people going wild with token consumption for things that aren't going to deliver outcomes for nCino and for our customers, but they have to understand that this is a safe space to go imagine a different way of working.

Jim Marous (21:18):

Well, it's interesting because I just interviewed Liz Wolverton from Synovus now as Pinnacle, and she said it's about leadership, about communication, it's about culture. And if those aren't aligned, if you don't have a North Star to where you're heading, you won't get there.

Sean Desmond (21:34):

That's right.

Jim Marous (21:34):

But even more importantly, she's dealing with AI at the same time of a merger at the same time as just the concern of employees going, "Do I even have a job tomorrow?"

Sean Desmond (21:43):

That's right.

Jim Marous (21:44):

And the reality is it gets down more than anything else (and you just said it) to the communication function. You've got to make sure everybody's buying in because if they're not pulling on the same side of the rope, they're pulling against each other, even if it's silent.

Jim Marous (21:57):

So, we talk about this integration of tools and it's great if it's all integrated and data is in place and everything else is in place. Your solutions that you're going out to marketplace for and what people are buying, are they buying what I'm going to call partial solutions? So, segmented solution, they have a challenge and you're saying, "I can bring you an AI solution that's going to answer that question without having to change the whole organisation."

Sean Desmond (22:20):

So, nCino is a platform, and by the way, we're humbled to have Synovus Pinnacle as a customer as well. So, we're leading through that change together. But nCino's vision has always been to be a platform.

Sean Desmond (22:23):

When I talk about the things that we do across onboarding, account opening, loan origination and portfolio monitoring, those are solutions that are all part of the same platform, underpinned by a foundation of data, and shared data across those motions and how those banks interact with those customers.

Sean Desmond (22:51):

And so, when you think about having data from 2,700 customers, doing the things that I talk about across commercial, consumer and mortgage lines of business, and you talk about being able to serve up those insights to then predict the next action for a banker or their customer, it's because the platform brings all that data together, and that data has been carefully curated over time.

Sean Desmond (23:15):

And by the way, this isn't just that you hear a lot of even some of the largest software companies in the world talking about access to data, and they're talking about maybe the hooks and the pipes into the data, and I've lived that world. I came from Informatica before I was at nCino, a horizontal data integration play.

Sean Desmond (23:32):

What we're talking about here is not access to data at nCino. We're talking about real, actual, real world, process-centric, curated for financial services, data that has a point of view on how capital flows through workflows and how risk is managed, and how decisions are made with respect to money that goes out the door. That's the data that we're talking about.

Sean Desmond (23:57):

And so, that's so differentiating from, listen, I had yet to talk to a bank CEO of any size or scale that says they want to automate their business on public cloud data. They want data that has an actual informed point of view and that's nCino data.

Jim Marous (24:12):

Well, and AI gives you the ability to use outside data to enhance this. Certainly, the commercial lending area, any lending decision, use data that was never accessible before, and actually build solutions that way.

Jim Marous (24:24):

So, you had a client up there on stage today that truly is embracing AI. There's no doubt about it from beginning to end himself and his company as well. What do you see as the difference between those organisations that are out there at the lead of the pack, and those that are maybe what we used to call in the banking world, fast followers, which is no longer applicable because the industry's working so fast. You can't be a fast follower, you're a follower.

Jim Marous (24:50):

So, what is the difference you're seeing between those? Is it all in the top or you see it mostly being in the leadership?

Sean Desmond (24:57):

And that's exactly why I connect so well. We were on stage together because I think you have to be a little bit candid, and you have to put yourself out there, and you have to be willing to go on the journey with your team. But you also have to just have the posture that while we talk about banking being this risk averse industry and we talk about regulation, all things, because every business has some risk, right?

Jim Marous (25:23):

Oh yeah.

Sean Desmond (25:24):

And I talk to bank CEOs every day, and I do hear a cohort of them say, "We don't want to be first. We want somebody else to figure it out, and then we'll jump in."

Jim Marous (25:35):

We said that about instant payments too. The same biggest organisation said, "We're ready to go, but we want to be the first out of the block." And the reality is it taught us very quickly between first and third and 50th, there's a very small ... it used to be one after another. Now, we're talking about hours, days, weeks, months between the fastest and the ones that are slower.

Sean Desmond (25:59):

And listen, we're all informed by our experiences. So, maybe that cohort is there for a reason. Maybe they've been burned before. Maybe they did have some sort of a regulatory-

Jim Marous (26:08):

We all have our reasons.

Sean Desmond (26:09):

And so, it does become cultural, but the reality is the folks that are willing to be first and get it right are going to be the ones that run out front and widen the gap. And I do believe in the other narrative in the market that we're all hearing, is that M&A is picking up and consolidation is real. And I believe that the folks who embrace AI are going to be the folks who are the big winners in the consolidation game.

Jim Marous (26:36):

Well, it's interesting. Over the last month, I've had four people I've interviewed reference in major consolidations down the road. Now, we've talked Sky has fallen, Chicken Little a long time. We've said, "We have too many brands, we have too many main banks, all this." But I think AI is going to really be the differentiator because you're either there or you're not, and the acquirers are going to have to be there, or else you’re not going to be able to get scale.

Jim Marous (27:00):

And those who say, "You know what, I can't afford this right now …" And it's not just size, it's still leadership because you got to have a very big regional bank that's just not going to be ready because they haven't gotten out of their mindset of what used to be banking, which could have been five days ago or five years ago.

Sean Desmond (27:18):

And it's not only AI. AI is the dominant narrative, but you look at the conversions of modernization that's happening. You think about AI, you think about the dual workforce, you think about digital currency and stablecoin and where that's all headed.

Sean Desmond (27:34):

And whether that takes three years or seven years, where is it all headed? You think about the banks that have the posture of embracing a new operating model, and imagining that there may be different rails that we run on in addition to agents that work next to the humans, all of it together really reinforces that the longer you wait, the further you fall behind.

Jim Marous (28:00):

So, you're an organisation, you've been going fast. What is the one thing that caught you by surprise, either positive or negative that you go, "Wow, I didn't see that as being part of this journey."

Sean Desmond (28:14):

The biggest surprise for me is that right now, there's not a really clean way to actually measure cost. And so, you do have to be careful there. And so, what you're seeing is even some of the hyperscalers who obviously are dominating the airwaves and what's going to happen with investment. And the hyperscalers have for a moment just hung back and said, "We just want our customers to use the technology. We don't want to dial up the pricing because we don't actually know what our cost basis is going to be in the long term."

Sean Desmond (28:53):

And so, when I think about the partnerships that we have with these folks, and then our ability to pass an experience to a customer that has an nCino cost as well as a pass through hyperscaler cost, we have to get that right. And we take on that ownership, we take on that accountability for our customers. So, when they leverage-

Jim Marous (29:11):

So, you're investing on their behalf until you know what the cost is. Don't know where the landing zone’s going to be.

Sean Desmond (29:16):

And we're measuring the cost. And so, right now, as we think about rolling out our digital partners and if we have a credit analyst agent that is delivering 60% efficiency on a credit review, it's drawing down on what we call intelligence units, what a lot of the industry calls tokens.

Sean Desmond (29:31):

And so, we're watching very carefully the correlation between those tokens and the outcome. And that's what I think most folks, they're not there yet. They're just actually looking to get the micro wins without connecting to the outcomes, and figure out the cost later. And that's not necessarily new news, but it was a surprise to me that there's a long way to go in the industry on understanding that cost basis.

Jim Marous (29:57):

So, finally, Sean, we used to say 5 to 10 years — that doesn't happen anymore. So, looking ahead, one year, maybe next year-

Sean Desmond (30:05):

Thank you for that.

Jim Marous (30:06):

Next year's nSight (yeah, I don't want to hold you accountable for that), what do you think is going to surprise bankers as to how agentic AI, AI in general, what's going to be out there you're going to go, “Wow, you maybe didn't see this coming?”

Sean Desmond (30:21):

I really appreciate because I do have a running joke internally with my team — I say somebody asked me, “Where's the company going to be in five years?” Let's just assume they asked me that question five years ago, I wouldn't have said anything we're doing right now.

Jim Marous (30:35):

Yeah, exactly. And you probably wouldn't put yourself in this position of doing a podcast exactly.

Sean Desmond (30:40):

I don't think it makes me wrong. So, one year from now, what I would tell you is we used to make this assumption on the amount of people and the amount of time it would take, let's just take an example to make a credit decision on a complex commercial loan at scale.

Sean Desmond (31:02):

And there was always this 7 to 10 people collaborating in that workflow, in and out: credit analysts, underwriter, loan officer, everybody in between up to the top of the house. I do really believe that because of AI for the first time, we can realistically with credibility, imagine what we call the path to one, where you could have a single human involved in adjudicating a complex commercial loan end to end. And as we compress the time, I think that's what gets lost often. The biggest time savings to me is if we minimize the handoffs.

Jim Marous (31:40):

Oh, gosh, yes. Which also minimizes the mistakes.

Sean Desmond (31:43):

It does. And the hardest project you do is the one you do over. I learned this in the deployment world. And so, the projects you typically do over is where there's a disconnect. And you think about in banking, all the people that process the … in software, product manager, engineer, QA testing.

Sean Desmond (32:04):

And so, the more AI compresses those roles and democratizes that less people can do more things, the less handoffs you have, the less risk you have of disconnects, and you just get the work done.

Jim Marous (32:17):

So, that's efficiency. Do you see this also possibly ending up where we're going to be offering larger lines of credit? So, more further down what would have been the traditional credit score mentality, both on the consumer and the commercial basis. Do you see credit being more available because we're going to have more information that's going to make us more secure in doing that B or C loan as they used to call it back in those days?

Sean Desmond (32:39):

I do. Listen, you're going to have to at some point trust the data. And I used to hear bank CEOs say, "We will only automate loans of a certain size, up to 50k." Now, that number’s going up, up, up.”

Jim Marous (32:58):

Because security's feeling better. And you have more information than five people could have ever pulled together on this company, and being able to also monitor that credit over time. So, it's not just the offering of a line, it's actually being able to monitor it on a minute-by-minute basis to avoid that risk. And eventually, that risk return thing is going to pay off.

Sean Desmond (33:18):

And it's auditable and it's traceable, and you can review it, and you're not dependent on a single human in the back office that is going to make all those decisions. You actually are having a human interact with the data, so I think that data is powerful.

Jim Marous (33:33):

So, as a takeaway for my audience, couple things. Number one, if you aren't using AI on an ongoing basis to learn what's possible, you have to do it. He did it. He jumped into the fire, said, "I'm going to get ..." And he's benefited from that.

[Music Playing]

Jim Marous (33:47):

Number two, don't see it as only a threat because the reality is as AI and humans work together, you're going to be offloading some of the things that used to be done that were repetitive and were not much fun to do, but enhance the intelligence, the working intelligence of the humans involved.

Jim Marous (34:02):

Number three: communicate, communicate, communicate. As an institution that’s really doubled down on AI, he has to communicate to his team going, "This is not a threat, this is an opportunity." And the last question we handled is while it's going to make some efficiencies, it’s going to eliminate some risk.

Jim Marous (34:20):

It's going to actually bring a bigger opportunity possibly to markets that we really had underserved in the past because we're going to feel more comfortable with the information we have at our fingertips.

Jim Marous (34:31):

Sean, thank you so much for your time. Really appreciate not only your insight, but also the insight you're sharing with all your clients at the event.

Sean Desmond (34:39):

Thanks for your time and for hosting, and I love your summary. Perfect call to action.

Jim Marous (34:44):

Appreciate it.

Jim Marous (34:47):

Thanks for listening to Banking Transformed, the winner three international awards for podcast excellence. If you enjoy what we're doing, we would really enjoy a positive review. Also, check out my recent articles in The Financial Brand, the research we're doing for the Digital Banking Report.

Jim Marous (35:02):

This has a production of Evergreen Podcasts, a special thank you to our senior producer, Leah Haslage; audio engineer, Chris Fafalios, and video producer, Will Pritts.