Whitney McDonald 0:05
Hello and welcome to The Buzz a bank automation news podcast. My name is Whitney McDonald and I’m the editor of bank automation News. Today is September 5 2023. Joining me to discuss AI use cases implementing the technology with safety and compliance in place. And a forward look to Predictive analytics is Dustin Hubbard, president of digital innovation provider archway software. Dustin has spent his career in tech and most recently served as the Chief Technology Officer at Washington federal bank before moving into FinTech join me in welcoming Dustin,Dustin Hubbard 0:35
Hi Whitney great to be on the podcast today. Thank you so much. I’m Dustin, the President and CTO of Archway Software. Simply put our choice of SaaS company. And we serve financial institutions looking to increase their revenue, their reach, and their relationships through digital transformation. We do this by providing world class products deployed on our enterprise grade architecture that may Nabal these banks and credit unions to succeed and what’s really become a highly competitive industry. I spent my entire career in tech spending 15 years at Microsoft, where I ran a lot of software teams and and products. Maybe the one most notable is the Snipping Tool, which is in every Windows SKU ship, that was actually a product I worked on when I was young in my career. I went on to help found an insurer tech company that was eventually sold to a fortune 300 insurance carrier. And then I was a CTO at Seattle regional bank, before I became the president of archway software. So that’s a little bit about my background.Whitney McDonald 1:46
Great. Well, thank you for being here. Of course, we’re here to talk about all things AI. Let’s kick things off by setting the scene here with AI today, of course, we want to get into the future look and where it’s headed. But let’s take a step back and talk through AI today, I was hoping you could talk through some real world examples of how AI is being used within finance today.
Dustin Hubbard 2:08
Sure, I’m so glad we’re talking about AI because not only is it on everyone’s mind, but it’s greatly misunderstood. And like any new technology, it’s disruptions can have positive and negative implications. So it’s important to think of use cases that aren’t going to give your compliance team a total heart attack. So as it relates to AI and finance, it mostly tends to fall into three buckets. Today, you have fraud detection, you’ve got virtual assistants, or chatbots. And more and more, you’re starting to see things around marketing and cross selling. So far to look in maybe a little bit deeper on, let’s say, chatbots, because that’s the one that I think is most visible to people in something that’s relatable. The virtual assistants our program to answer common questions, is designed to obviously improved self help, but also reduces the burden on the contact centers. That’s why businesses are interested in them. But as we all know, these are not all created equally, and how a customer feels about chatbots. And virtual assistants vary a lot. And the reason is, because a lot of these actually aren’t using AI at all. So a lot of the early chat bots are effectively decision trees, right? You’re answering a question that you know, the person is going to ask with a canned response. And those examples, typically, we don’t answer the question correctly, or you haven’t predicted what the question is going to be the bots like, I have no idea what you’re talking about. Please rephrase. And then the customer gets frustrated. But the chatbots using actual AI models underneath of them are becoming much more human like in their experience, which makes a dialogue between the bot and the human feel much more natural, and also is far less likely to get stumped with questions. So I think that’s a different kind of a differentiating component of how Chatbot is evolving. Now, the one use case I didn’t mentioned underwriting, and that’s largely because of concerns and bias modeling. So regulations around adverse lending is a really serious thing for financial institutions. And that’s a serious issue if they’re humans making underwriting decisions and a serious issue, if a bot or an AI system is making those decisions. So typically, I think even though underwriting is a use case, in AI for financial institutions, you’re not seeing it applied very much because of the concern that the the model could be biased in that actually probably complained to the heaviest regulatory scrutiny.
Whitney McDonald 4:44
Now, you mentioned chatbots, you mentioned decisioning. Of course, AI, even just in the past couple of months has come a really long way. And it seems like it’s changing almost daily. Now. How can a financial institution approach implementing this new tool technology like they’ve implemented tech in the past?
Dustin Hubbard 5:03
Yeah, sure. FIS know that AI is going to become a more and more critical part of their underlying bank operations. I think that’s a certainty. But similar to cloud transformation, there’s a lot of concerns with risk, there’s a lot of concerns with in house expertise and knowledge on how to do it, how to deploy it. In fact, it’s hard to believe that AWS has been around for nearly 20 years already. Yet, banks are still relatively early in their overall cloud transformation, compared to a lot of the other industries. But the one difference also between cloud and AI transformation is speed is a necessity with AI for banks. So they need to have a sense of urgency. And the reason is cloud transformation was really about modernizing the hardware, making it more scalable, improve resiliency, better, maybe security, but it wasn’t customer pressing, your clients weren’t more impressed with the bank, because they’re on the cloud versus in the data center. So banks, I think, have a little bit more time to work through the rest of the mechanics of doing the deployment. With the AI, it’s going to be detrimental to their business if they don’t start adapting quicker. And so when they think about how to start applying it, I think they need to think a little bit more around who are the right partners and providers that are going to help them do the implementation, AI has become more of an appliance, meaning you buy the AI off the shelf, and a plug it into your business model, as opposed to like designing your own AI models from scratch, right? Banks probably shouldn’t be in that business. It’s too complicated. And so I think that that’s the biggest difference. They need to find who they’re going to work with, they need to find the use cases that they want to start with. And it’s a classic crawl, walk run approach.
Whitney McDonald 6:49
Wondering if you can expand on one thing there, which is you mentioned it could be detrimental to your business. If you don’t implement AI, what could that look like if a financial institution does kind of take too much of a hands off approach or keeps it too much at an arm’s length?
Dustin Hubbard 7:05
Well, I think there’s a couple of issues. One is AI is going to be a game changer in terms of banks, operational scalability. So as AI starts to change the economics of banks, which is how many people they need, how much operations can be automated banks that don’t apply AI are going to start having their margin squeezed, I think a lot more than banks are actually effectively using it. The second part is banks are using AI are going to be able to more proactively serve their clients. And so as a customer, if I start looking at Bank, a, who waits for a crisis to occur, and I walk in, because I tell them, I’ve got a problem, and Baby B who tells me, you’re gonna have a problem, let’s do something proactive about it, people are gonna naturally think gravitate towards those banks. So those are customer retention component there as well.
Whitney McDonald 7:56
Now, we mentioned a little bit about compliance and how to implement this. If we could spend a little bit of time here on how financial institutions can keep up with AI implement in a safe and compliant way. How would you? How would you say that if I should approach that, and then maybe an example of a bank that might be doing this? Well.
Dustin Hubbard 8:18
For starters, FIS can’t let the perceived risk of AI causing action at that’s the big takeaway. If they do their input, the longevity of their business, I think in serious jeopardy. So one of my favorite use cases is around Lafayette bank, and how they transformed their contact center through the use of conversational AI last year, through COVID, in particular, staffing shortages really magnified the call center experience with long wait times customer frustration, people that really don’t know how to serve you. And this is really across all industries and banking was no exception. So one solution wasn’t to hire more agents, which I call the kind of brute force method just hire more people to handle the calls. It was really to improve self service and call deflection through the use of conversational AI. So in six months, they replaced their entire call center stack, inserted conversational AI at the very top of the call, and included voice biometrics, which meant the customer would not only reduce their fraud, meaning the bank knew that it was the person on the other end of the phone by meant that customers could start doing self servicing. Like how much is you know what, please read me my most recent transactions, and the data speaks for itself. WAPA just posted their highest net promoter score in their history 57 the industry average for banking is 35 years ago, wall fence was 17. And when people are voice enrolled when they have the voice biometrics enrollment, they’re seeing only 5% of the time are those clients asking the bar Up to talk to an agent 55% of the time, if they’re not voice biometrics enrolled, they’re asking to talk to an agent. So clearly, when they’re enrolling in voice biometrics, they’re able to do self servicing through conversational AI, it’s actually improving the client experience, but also reducing cost and burden. The point is, AI can be safe and compliant, really, by picking use cases that are well understood, and also already proven by RFIs. They don’t all have to be riddled with like massive risk.
Whitney McDonald 10:32
Now, thanks for sharing that example. I think it’s important what you just mentioned that you don’t necessarily have to pick the riskiest ones in order to implement AI. I think that brings us into the question that we’ve been excited to get into, which is, where is AI headed? What are those future use cases, whether it be short term or long term, wondering if you can talk us through how you’ve been exploring AI and what you see for it in the future?
Dustin Hubbard 10:58
Definitely. And if people follow me on LinkedIn, they’ll know I’m certain AI is can become the most disruptive innovation of the century. For banking, that’s because it has the capacity to dramatically impact every banking function, from origination, to fraud, to lending, to servicing, and eventually to actually predicting. So let’s imagine for a minute what banking might look like in 2030. That’s one of my favorite things to do as kind of a technologist and how AI would actually pull that change. Today, nobody’s paying attention to your finances, but you there’s, you’re the only one who logs in your loved one knows where your money is. And if there’s a problem, you’re the one who’s got to sort it out. But in 2030, actually think AI is going to be doing that automatically with you. So it will know when you’re going to be short on cash between pay periods. Maybe before you do, it will know if you’re going to default fall on a mortgage before you realize that you have a financial crisis, three months on the horizon. So maybe an analogy I could use is to think about how we dealt with weather before the use of satellites. Basically, before we have satellites, we knew if storm occurred when it landed on our front doorstep. And there was no forewarning whatsoever. But satellites completely changed that because it gave us the ability to see weather before it hit us, thus allowing us to prepare to evacuate, to do the things we needed to do to protect ourselves. And that helps minimize the loss. So for banking, I see AI solving that problem for finances. The way satellites help solve that problem for whether it’s predictive, it’s preparedness. It’s not just reactive. So the point is, many technical innovations over the years have disrupted banking from debit cards, mobile apps, peer to peer payment systems. Banking hasn’t gone away, but it has changed and AIS can change it again. But I think vastly more profound ways than probably all those other ones combined. So Fy is really need to have a five year proactive AI plan. They need to work to implement it and refine it and basically make sure they don’t become the next blockbuster. To me that’s the biggest thing that can change the next five to 10 years.
Whitney McDonald 13:26
You’ve been listening to the buzz, a bank automation news podcast, please follow us on LinkedIn. And as a reminder, you can rate this podcast on your platform of choice. Thank you for your time and be sure to visit us at Bank automation news.com For more automation news
Transcribed by https://otter.ai
