Altaone FCU AI Serve Members Win Tomorrow

Thank you all for joining. We’re excited to be here. We have a wonderful partner, institution we’ve been working with, for several years, AltaOne joining us. And today’s discussion is is centered around how they’re using AI to serve their members for today, and how to win tomorrow.

So we’ve got a couple of things that we’re gonna cover here today as we move into our agenda slide.

First, about Interface AI. Most of you may have already interacted with us, Maybe you’re existing customers, partners of ours, but some maybe this is your first exposure. So who are we? What are we doing?

What what are we trying to solve? So we’re gonna go through a quick introduction to the company here. We’re gonna do a quick demo of the Interface AI of voice, and this is a a primary topic today as our our presenters, and we go through our panel, our case study. So voice is gonna be critical to the conversation.

So we’re gonna show you what that looks like just so you have some context as we get into it. Next, we’re gonna get specific. We’ve got, I’m gonna introduce him here in just a second, a wonderful presenter here with us today from AltaOne who’s gonna talk about, you know, how they’re using this and and what what tools and leverage they’re seeing from it.

We’re gonna also give you a little glimpse to what the future is. This conversation, again, is is so hot of a topic in today’s industry, and there’s people thinking about AI differently. And not only from an opportunity standpoint, but from a risk standpoint. And and what are those conversation points, and and where are people thinking things are going today?

So we’re gonna get into a a couple demos, Sphere and Nexus. These are coming products here for interface. And then we save some time at the end here for q and a. So please use the q and a function.

We will be monitoring that on the side. We’ve got the wonderful PC on here who gonna help answer some that he can. If not, we’ll try and take some of those live at the end.

Enough of me. Let me introduce oh, sorry. I forgot. We gotta get your fingers moving first and your brain’s working.

Thinking about today’s topics. We have some questions that we wanna pose to you, trying to get you in the right frame of mind here. So how would you describe your current use of AI, member customer experience? There’s a lot of tools out there.

So what are you using? How is it working? Putting you in that frame of mind.

We’re gonna give a couple seconds here, maybe a minute or two to answer all of these questions.

I’ll jump to the second question here in just a second.

Give you a chance to think about it.

Question two. What is your primary goal for AI in the next twelve to twenty four months? This is a tricky one. Things are moving so fast. That answer is changing almost daily, I bet, for some of you.

Question three.

What is your biggest concern about using AI today? There’s a lot of noise out there. There’s a lot of conversations happening around this.

What’s the perception that we’re seeing today?

Question four.

How familiar do you feel with terms like generative AI versus agentic AI?

This is a knowledge test. You guys been doing your presearch before you jumped on today’s webinar.

And last but not least, I would not be good at what I do unless I asked who would like to see a custom demo of an AI solution. So please go ahead and indicate so here at the bottom. Again, no pressure.

Give another thirty seconds or so. I know there was four or five questions. The fifth being really important also. Take a second. Think about it, and we’ll go ahead and take a look what that those results are.

K. I’m thinking since we’ve got such a probably amazing group here this morning, we’re we’re ready to go. There we go. Okay.

So just exploring no live use was about fourteen percent of our first question. So the again, that first question being, how would you describe your current use? The biggest, live in one or two areas. Interesting.

Contact center collections.

The smallest sample size, advanced, working towards a or autonomous use case.

Common answer. I think a lot of the conversations we’re having, and I I’m glad you guys are here today because I think you’re gonna see where the world’s heading, and a lot of how people are thinking about this. So this is very insightful.

What is your primary goal for AI in the next twelve to twenty four months? So biggest answer, almost a tie there between improved self-service and satisfaction and reduced operating costs, improved efficiency. Obviously, very, you know, critical components as we serve consumers, right, especially in the financial industry. So makes a ton of sense there.

Reduce fraud and risk coming in at the least. Little surprising, but good indicator for us. I think there’s gonna be some interesting conversation around here as well because this is obviously going into that risk bucket.

What is your biggest concern about using AI today?

So leading answer, cost, complexity, and change management.

Interestingly enough, there there is a lot of ways to look at that. And so, you know, at certain points, it may be appropriate to have conversations. This can be simpler than you think. And a lot of, you know, solutions out there can be complex, but agentic modeling and what you might learn today is solving for a lot of complexity.

That’s the the almost a a main purpose, right, of it or main pillar. So thank you for your answer there. And then I’ll move quickly here because I know you guys are more interested in hearing from our other presenters. How familiar do you feel like with terms of generative AI versus AgenTic?

Somewhat and comfortable are our middle ground. So right in the middle there. So we got people with some exposure here, maybe not fully embedded into to every all the terminology, but hopefully today helps solve that. And then great responses here.

So I’m gonna go ahead and close out that poll. K. Let’s go ahead and move forward to our our presenters. I, myself, I’ll go ahead and go first, and I’m gonna be quiet here and and turn it over to the real geniuses.

Mike Lazavito, VP of sales here at Interface AI. I am your host for today. I get to ask some wonderful questions and challenge our other two presenters.

First, I’ll introduce Sri, and some of you probably known Sri from industry or know him as our CEO here at Interface. Obviously, a a brilliant mind in the space. Been challenging the space for a while. Excited for his input. And also joining us from AltaOne, and a a lot of appreciation for you, Phil. Phil Fowler, chief experience officer. Really excited for his input, the way he views this space today, where this space is going, and a lot of what a lot of our institutions should be considering because the world does change now almost on a daily basis.

With that, I’m gonna turn it over to Sri to take us through interface. Sri, thank you.

Thank you, Michael. Thanks, everyone, joining.

Hope you’re all having a good good day, good morning.

You know, Interface AI as a company been around for the longest standing AI partner for in the industry. We’ve been around since twenty fifteen.

We’re headquarter here in San Francisco, and we’re about hundred and fifty employees. We’re also a QSO.

You know, what we do is we are completely reinventing banking self-service in the age of AI. Right? So we call it intelligent banking or what we call bank GPT. This is purpose built for credit unions and community banks. We do not serve any other industry or any other kind of business.

You know, I personally come from a a credit union family myself.

You know, my father grew a Credit Union from from scratch to a couple of billion, and I am also ex Microsoft, part of the AI research team. I combined those experiences to build interface AI.

You know, it’s it it it couldn’t be more exciting with all the capabilities of AI to, you know, use them to kind of really help the, you know, community banking space.

Very passionate about it.

My cofounder, Bruce Kim, was a CTO. He comes with over twenty five years of experience building really complex, large enterprise scale systems.

You know, he previously built a billing software world’s first billing software, which grew up grew to fifty million plus in revenues.

We got together on this journey to really reinvent community banking using AI. We have been featured in Gartner, Kuna, Allied as the top AI vendor for the industry, including winning the best AI award from American bankers recently.

We today serve over a hundred financial institutions, you know, of various asset range from half a billion to all the way, sort of tens of billions or thousand assets, you know, quite a bit of range.

Yeah. So if you can switch to next slide, Jack, real quick.

So I know some of you already mentioned about the knowing the difference between generative AI, agent AI. I’m gonna keep it brief.

This is this slide is probably the simplest way for you to understand these you know, the terminologies that everyone use and really what they mean in the context of self member and employee self-service. Now, of course, these technologies can be used in underwriting, fraud in many, many places, but we try to simplify and make it easy for you to understand what these terminologies mean in the context of employee and member self-service. Right? So if you if you look at it, like, you know, because we’ve been in this space for eleven years, you know, we have been through all of these technology waves.

Right? So if you see on the left side NLP bots, this is where we started. You know, literally, in the context of self-service, when you use this NLP bots, you know, the bots were, you know, it’s you know, pretty poor in understanding the member request when they type in the question or they call in or or it’s an employee request, and the response was highly scripted. You know, those were the days where we work with some some of the customers who joined here.

We work with you on literally scripting the responses together, very manual, and all of this. It was but we still you know, some of the visionaries, some of our customers still leverage this technology to gain significant value with automating lot of customer request and all of this. But then came the generative AI, where in generative AI, the AI become really good in understanding the request. So you no longer had to, manage specific keywords for it to really understand the member request or an employee request.

But the responses were still scripted because it was, you know, hallucinating a lot in the early early days of generative AI. So, you know, you were not training the keywords, in understanding member request anymore, but but you’re still doing all the responses were manually scripted. That that’s where we were about three years back. Now, you know, the later came fully generative. As a generative AI technology become a little better, we were able to use that in response as well. Right? So, of course, the request member request, employee request, AI got better in understanding, but the responses now were also fully generated by AI, but it was good enough to do it for the informational questions as in behind the you know, in front of the pin kind of questions.

You know? And then came the agent AI where it’s, you know, far superior in understanding member requests. It almost feels like you’re talking to a human being. And it also not only generates the responses in front of the pinned responses autonomously, but it also has a ability to ask you a follow-up question and, you know, not, you know, throwing bunch of text from your website or document like the generative AI used to do.

But the agent AI ask you a little bit clarifying questions and finds the right information for you on the informational request. But on the transactional one, it is so powerful. It has ability to, almost follow the rules like a human being and then, you know, call the right APIs to fulfill a a fund transfer or a loan payment. And it goes one step beyond that, like, you know, which is kind of really important in our industry.

The agent AI doesn’t stop in fulfilling request that actually have APIs.

It actually goes one step beyond that. It actually goes ahead fulfill request as long even using your the desktop and browser tools that your employees otherwise use to fulfill the member request, it has ability to automatically navigate, your desktop tools and browser tools to actually, fulfill a member request without a human intervention or with a limited human intervention. Right? So that’s what really agentic means.

Right? So which is, it has ability to, you know, go deeper and reason, of what you’re asking and ask your follow-up questions instead of dumping you a bunch of data and go call your APIs as needed. If there’s no APIs, it has ability to navigate a tool and complete the request. And, you know, and and that’s, the basic level of agent AI.

And then kind of there’s more advanced, which is agent AI can also be very consultative. Right? If someone comes with a problem, it has ability to, you know, reason through the problem. Okay.

I’m having, you know, issues with, I would like to consolidate my debt and resolve my, you know, credit card debt. And it has a ability to con consultatively offer solutions, guide them through the process, and things like that. So that’s where agentic AI is current technology.

It’s, you know, it’s still new for banking. Right? Because, you know, there are there are other industries who always move faster because there is less risk in them. You know, if if, you know, let’s say, McDonald’s, you’re driving through and they can take a chance on using agent to get faster than all of us sometimes.

And because if if you if your burger if you’re ordering a burger in the drive through and you miss out the ingredient, no one cares. But, you know, in banking, if if you end up sending money to a wrong person and things like that, it it will be significantly catastrophic. Right? So which which is the reason agentic AI is not widespread adoption in banking yet, but it is getting there.

And we are pioneering the ways. And we have one of the visionaries here, which is Phil, who worked with us closely to make that happen, and he’s gonna talk about the journey of him his company moving along with us from NLP bots to agentic AI. They have been with us all along.

So he’s gonna take us through some of his experience. And, you know, so there is visionaries who are already adopting this and working closely with the partners like us who specialize in this. Right? So we’ll go through that. Can you go to the next slide, Jack?

So this is you know, with agent AI coming to a reality, this is where things would change. Like, the I don’t think it’s just a a change in your member experience or something like that. It’s a fundamental change in your self-service strategy. In many ways, it’s a fundamental change in your banking strategy.

Right? So it’s it’s pretty significant. Right? So if you look at the kind of the two paths you could think about, we can’t continue to go deeper in unlocking layers of agent to get impact, but this is kind of the first layer.

Right? So if you look at, you know, how, agent to get it could be used. A typical partner you you may be talking to, you know, they have agent to get deployed on voice and chat. They say, oh, we’re gonna automate your chat and voice support.

But, you know, that’s twenty, thirty percent of our members really interacting on voice and chat. The sixty, seventy percent of our members are using online mobile banking who are still getting a a non agentic experience.

So our approach is a little different. Right? So our approach is we are actually also building a AI native online mobile banking. So your member experience is is not intelligent in chat and dumb in online mobile banking.

It is gonna be intelligent across the board. Right? So that doesn’t mean it’s just a a a you know, adding a chatbot in online mobile banking is far beyond that. Right?

So we’ll we’ll talk about that, today as well. And Phil has quite a huge vision on that. We are partnering with him to, make this happen, in terms of deploying an AI native online world banking. Is has a a lot of great insights to share.

So if you switch to next slide, you know, that’s at the member experience. But AgendaGeya goes deeper. Right? So which is it also disrupts and reinvents your employee experience.

You know, with AI ability ability to being able to automatically navigate tools, the desktop and browser tools, Imagine your frontline employee life. They have three different monitors. They have to learn to operate across fifteen different applications.

And for every member request, they’re probably going through a couple of different application to fulfill them. Right?

So you know? But with the agent AI, you would see them going to a one screen and which we’re gonna show you today. The employees has ability to come in when AI gets stuck and and and, you know, just unblock the AI and let it do another ten steps. So, you know, it’s fundamentally re fundamentally rearchitects the self-service stack because your self-service is now not only agentic and intelligent, but it also has ability to use the staff per and personnel in the financial institution in in the loop to resolve any request.

Right? So which means it fundamentally changes your employee experience too. Right? So, everything geared towards fulfilling member request, you know, proactively or reactively.

So and then we kind of go one layer deeper, how agent again continues to, change this. If if you go to next slide here real quick. So this is the next level deeper. So if you look at where we are as a as a company and, you know, which is kind of really at the cutting edge of agent AI, and then we’re really pushing the horizon, the the yellow says what are the things we are kind of doing currently.

It is you know? And then the and the blue, talks about things that we’ll do very soon as AI continues to get better. And and these are all, it’s kind of we are riding the way of AI and becoming more intelligent, and we all we wanna take you through that journey with us. Right?

So if you look at Yellow, you know, where we are, which is fully maximizing the use of current AI capabilities to reinvent member and employee self-service, We have this one AI brain that you could deploy across multiple channels. There’s voice chat, you know, or text message, that’s kind of table stakes. Right? And then the AI that the agent of ours today also has a ability to not just fulfill a member a member request or, you know, when they ask something or they ask for a fund transfer, of course, it’ll fulfill that.

But today, it also already has ability to do an upsell cross sell. Like, it says, hey. You know what? I helped you do the fund transfer.

By the way, you know, it seems like you you got a a a car recently. Do you wanna sign up for an extended warranty? Right? So it it has ability to take, you know, member request.

It’s like it’s like Google search. It’s like Bing search where I worked. If you go to Google search and type in a query, it it shows you what you asked for and also shows you some, you know, ads that tells you something relevant to you.

Similar to that, when members are coming in and asking you to do something, you’re using those interactions as an opportunity to upsell something or product a financial insight that really helpful for them. And then we also use a computer use agent, is basically a a more technical word for intelligent RPA.

It’s where it has ability to navigate the screens and autonomously fulfill a member request, you know, when there is no APIs. Right? And then with Phil and the team, are working on building out the AI native online mobile banking, which replaces the traditional online mobile banking. These are things in progress.

But where this could go as AI continues to get better is, you know, very soon with this AI, you know, they they should be able to add multiple financial institution without even using Plaid or MX. And then when they add, the AI should be able to use the bureau data and all the data from multiple relationship they have to, drive proactive financial insights and things like that to, help you make, their primary financial institution as well as help resolve their, you know, financial goals or achieve their financial goals. Right? So that’s kind of where it it will go very soon.

You know, just like how we are driving cars today without really, you know, having to drive, you know, you’re going to get, AI that can autonomously drive your financial goals for your members, which truly helps us to live up to our mission as credit unions and community banks, which is driving financial well-being for our members. And that is gonna get very easy very soon. Right? So it’s an exciting future.

And if we do end up solving autonomous banking, you know, you could really promise your members, you know, sort of basic income. Right? So so that that’s that’s a very much possible with how close we are associated to AI and how fast it is evolving. These are all very feasible goals in the next three to five years.

Right? So, hopefully, that gives you an idea how agentic AI is really transforming. It starts with member self-service because once you make that intelligent, that becomes the window to everything else the AI could do really well. Right?

So, and and, you know, hopefully, it kind of gives you multiple layered deep in terms of how it really reinvents banking and how we kind of how vision, how we’re shaping that. Right? So if you switch to next slide, Jack, real quick.

These are our solutions broadly today.

Doing this for eleven years, gotten a chance for us to build quite a bit of depth in these, you know, voice AI, chat AI, employee AI. These these products, we by far, the most deployed AI solution in the industry is from us because given the fact we’ve been around for eleven years, we have, by far, the most number of AI solutions live with our customers. Like, you know and and as the name indicates, the voice AI, you know, takes member request and automates it, automates the calls, and and Chat AI does the same for chat.

But they also have ability to upsell cross selling the conversation and drive new revenues. Employee AIs for employees to find help them with the knowledge finding access to knowledge and taking actions.

And then we have a few other newer solutions, which is a collection and the agentic online mobile banking and the contact center. We see some of the demos, how we are partnering with Phil to build some of these new solutions as well.

So let’s go to next slide, Chad.

And we have quite a number of integrations. We have built through eleven years of doing this.

You name a system, we most likely work with it. But very soon with the with the, you know, computer use agent, like the intelligent RPA we’re building, integration really doesn’t matter. You know, AI should be able to work on any system for for that matter to complete any task.

So that’s that’s where the world is heading, here really fast.

Great. So, Jack, if we can, move to the next slide. We’re gonna take a minute to show you our voice AI, which is what Phil and team used today.

And we’re gonna give you a quick demonstration of that before we get into the journey of Phil so you have context on what product they use, and you can easily follow the discussion thereafter. So we’re gonna show you that product here real quick. I’m gonna ask my colleague here, Prasunjit, PC, to help me out.

Perfect. Hey. Thank you so much.

And let’s straight away see a demo of how members can interact with the AI agent on a phone line. So I will share my mobile phone.

Show of thumbs. I hope you’re all able to see my mobile phone. Here, I’ll role play as a member, make a phone call into the voice line, and let’s see how am I able to have a conversation with the AI agent.

Hello. I am Alex. Welcome to AltaOne Federal Credit Union. Please note that this call may be recorded for quality assurance and training purposes. Your privacy is important to us.

You can ask for my help with anything. For example, you can say account balance, deposit status, transfer funds, transaction history, or loan payment. What can I help you with today?

Hey. Can you tell me more about the Veterans Banking Program?

One second. I will check.

AltaOne is proud to be part of the veterans benefits banking program, which helps veterans, beneficiaries, survivors, and caregivers access essential banking services and receive VA benefits securely. This includes personal banking, loans, VA home loans, online banking, financial education, and fraud alerts.

I can text you these details if you’d like.

Yes, please. Send me the text message.

Okay. One moment.

The information has been sent to your phone via SMS. Can I help you with anything else today?

Yeah. I was also wondering, do you have any accounts for children?

Let me find that information.

AltaOne offers the Cube Club savings account for kids ages zero to twelve.

It’s a great way to teach them about saving, managing money, and setting goals. I can text you the details if you’d like. Should I send that to your phone?

Actually, that’s okay. But can you tell me my checking account balance?

Working on that request now. Would you like to use device biometrics for authentication?

Yes, please. Go ahead.

Great. One moment.

I have sent you the text with the secure link to authenticate.

Please provide consent to continue with device biometrics login.

Thanks. You’re authenticated. Let’s continue. Your business checking account ending with seven eight nine zero has an available balance of one thousand dollars and a current balance of one thousand dollars. Would you like to check the balance of another account?

Actually, no.

Understood.

Based on your balances, you may qualify for a higher yield or premium account option that reduces fees and increases earnings. With your combined balances, you could qualify for a premium checking account that waives monthly fees, earns higher interest, and includes perks like free wire transfers. Would you like to hear about the premium account options available to you?

Yes. Tell me more.

Sure. One moment.

Great. Our premium checking includes higher interest, waived monthly fees, free wire transfers, and overdraft protection. Since you already bank with us, the upgrade is quick. Would you like me to send you a text with more details?

No. That’ll be all. Thank you.

No problem. If you ever want to explore it later, just give us a call or check our website.

So here, you heard how easy it was for members to have a conversation with the AI agent. And during the conversation, the member was able ask informational questions about the different kinds of account that are available at AltaOne. And then the member was also able to ask more specific information related to their account itself.

Best part, before even providing the information for the account, the AI agent first ensure that it is able to authenticate the member’s identity by using the device biometrics. So that is something which is unique to Interface, and we essentially use device biometrics, is the fingerprints or the face ID reader, which is already available on the member’s mobile phone to help unlock the mobile phone. That same thing can be used to authenticate when members are trying to ask for any financial questions. And our AI agents are intelligent enough to step up or step down the authentication depending upon the severity or the criticality of the financial transactions.

Knowing just an account balance, the AI agent is able to use the device biometrics. However, if the member wanted to make a money transfer or wanted to do a wire, at that point of time, the AI agent would layer in additional authentication models over and above the device biometrics.

And really using the device biometrics, it is essentially able to ensure that we are reducing the chances of unauthorized access and also any fraudulent or spoofing activities on the member’s behalf.

I have seen a few questions, and I can take a couple of questions live here.

Oh, I I see that somebody actually asked me to order a pizza. I didn’t see it, but other otherwise, I would have loved to order a pizza, and you would have noticed that our AI agents are fully grounded in the knowledge that it is being provided access to. In this case, the only knowledge that the AI agent has access to is pertaining to Alta one FCU. So which means if somebody were to ask the agent to order a pizza, it would refer to that request as invalid and essentially mention that is not part of the knowledge base.

In terms of the biometrics validation, the setup is done the first time a member calls in, and then the member would be given an option to enroll in the device biometrics, and a unique link would be sent to the member’s mobile phone. Once the member registers for the device biometrics, after that, every subsequent call whenever the member has a conversation and if they want to ask any financial related questions, the device biometrics would kick in first.

I I also saw during the poll that a number of the attendees today, you wanted to book a custom demo. So I have the QR code displayed on my screen. Please take a picture of this QR code, open up your cameras, and this will essentially take you to our landing page where you can book a custom demo.

And we would be happy to give you a demo talking about all the different features that Interface AI’s agents both over chat as well as over voice can be used to improve the member experience significantly.

With that, I would love to pass over the control back to Michael.

Michael, over to you so that you can then start the main part of the event today, which is the discussion with Phil.

Absolutely. I think, Jack, you have controls. Correct?

Yep. There he is.

Thank you, PC. Wonderful explanation. And and I’m excited. This is we’re gonna get into the fun stuff now.

But first, a little bit about AltaOne. I always love doing the due diligence before I get to speak with a lot of the financial institution partners we work with. The the history, the context in which they start is just so important, and and AltaOne, like many others, started as a small cash drawer on a on a naval ordnance test station back in nineteen forty seven in China Lake, California. Less than three hundred dollars, now about a billion dollars in assets under management.

An impressive, an impressive history, and where they are at today and where they’re going will no doubtedly continue to grow and impact their consumers and members in such a positive way.

So our first question, I’m gonna start with Phil. Put him in the hot seat quick here. Phil, how is AltaOne’s business strategy delivered more effectively within with AI in general?

Sure. Absolutely. Before I actually can answer that, I need to give you a little bit of background on how we got to where we’re at. About two and a half years ago, I met with Stephanie Sievers, who’s our CEO.

Excuse me.

I’m trying to decide, you know, if we were a good fit to work with one another, and she said something that really resonated with me.

And she said to me, we should be a fintech offering financial products and services, not a credit union offering technical solutions.

I went, wow. That’s really it really resonates. That is a extraordinarily visionary statement for a CEO to offer that. She doesn’t mean that we should start, you know, writing code. What she really means is we need to think like a fintech. And fintechs are always thinking of what is the next future iteration of something coming to solve a a today problem.

And Stephanie was talking with one of our information security analysts, And she asked him, hey. What are you using for, how are you using AI in your area?

And he kinda said, well, we kind of use it this way. I don’t really use it much a lot myself. And her comment to him was, if you’re not using AI, you’re bringing a knife to a gunfight, which also has now sort of become our mantra among the credit union. She really is the the visionary behind bringing, AI to the forefront here. In fact, I would even say that ChatGPT was launched November thirtieth of twenty twenty two.

She brought AI to her team in August of twenty twenty two and said, this is where we need to be focusing for the future. So she’s been really pushing this stuff for for quite a while. So when I started with her, the first challenge that we had is that we had high watermark sometimes in the call center for twenty minutes. Our members would call in, you know, if we just did a mass reissue of credit cards or we had an issue going on or on a big payroll day.

Sometimes Neppers would wait twenty minutes to get to an agent. And she says, we’ve gotta fix that.

And she thought that AI would be a great model, a great tool to sit in front of the contact center and try to answer some of those easy calls.

Some of the other things that we use I’m not gonna cover all of these just for the sake of time. But for efficiency, we use Copilot here at AltaOne, and you can have Copilot come in in the morning and summarize your email. What’s my inbox look like today, what are the key things I need to look at?

Can you help me reply to that email? Here’s what I wanna say and hit send. I need to create a memo. Can you help me create a memo? Here’s the the things that I wanna create, and then it’ll gather all the information from every context that it can find, create a memo for me to review and look at.

We’ve got an analyst, a financial analyst that he says, you know, I’ve never had anybody report to me, but I now have five Agencik employees that that work for me.

And I give them all kinds of commands and data to go analyze and bring back, and it has taken things I would work weeks on to where I’m able to get stuff done in hours.

Our cybersecurity team has deployed a solution that looks at all of the inbound emails looking using AI to determine if it is a an ID harvesting email looking to try to take over that account. If it is, it just does not deliver that to the to the inbox.

But we figured while it’s in there doing the analysis, can you determine if it’s a marketing email? If it is, let’s put that in a different folder so we don’t flood our inboxes with, you know, the eighty or ninety marketing mails that we get every day. That has been a fantastic, unbelievable time saver for, executive teams and others in the organization.

Policy drafting is something that’s very helpful. When you sit down, we gotta write a new policy for this.

And we’re trying to figure out, you know, how do we write that? I don’t I don’t really know that much about it. Well, we we query AI, and it helps us do those kinds of things.

Next up on our thing that we wanna look at is auto using AI to do loan decisioning. We have some automated loan decisioning, but at the end, it does never makes a it never declines anything. If it gets to a point where it can’t decision it, it goes to an employee who then has to make a human decision. We like to help that, by adding AI loan decisioning.

On the next slide, we actually talk about the branch planning, which is why I skipped that one.

We were looking in one of our markets to open the branch two hours later on Friday and close it on Saturday. We thought that would be a great idea, better serve our our markets. People could take care of their business on their way home. We were open later. And so we asked AI, hey, AI. We’re getting ready to close this branch on Saturday. Keep it open till seven o’clock on Friday night.

What is your input on that?

It actually came back and said, well, I’ve analyzed excuse me. I have a frog this morning.

We’ve actually analyzed the traffic patterns of the members in your area, and it indicates that most of them go home Friday night. And they they come out Saturday morning to do their business. So if you close that branch Saturday and stay open Friday, it will actually hurt members who rely on your branch the most. And we’re just like, wow.

Holy cow. We would have never known that. And without engaging AI to help us look at things that we didn’t know we had even had access to traffic patterns in the in our community. Wow.

It was it was it was pretty amazing.

So we’ve we’ve we’ve definitely, been able to leverage that that going forward there.

Thank you, Phil. So insightful. And I think one of the concepts you brought up going back to, like, how, what, why. Like, that’s as simple as it can be as a start when you’re using AI. Right? Like, how do I do this? Why would I do this?

Sri, I’m excited to hear what you have on this. What are the executives you are talking to say about using AI in their business strategy?

You know, I I think it’s it’s about really figuring out what’s the DNA of the organization. Right? Like and the the the strategy, of using AI and and impacting your business kinda changes a little bit based on the DNA. We talked to a lot of financial institution, like, have different kind of DNA.

Right? Like, one who are, you know, quite visionary. Like, they’re they they are all in on AI, and and then to all the all the way, the other end of the spectrum is they they wanna be cautious. They wanna start slow and things like it really starts with kind of understanding their DNA.

I I I think the when you when you I’m also assuming the folks on the call are are a different range of the spectrum. Right? So there is just so many ways to crawl, walk, and run.

You know? And if you’re ready for to be more aggressive, you can actually leapfrog quite a bit. And I’ll take you through a a couple of examples. Right?

Like, you know, if you’re on the left end of the spectrum, which is, hey. We’re we wanna we wanna take a cautious approach. Right? So what you could do, and then if you’re on the on the right end, what you could do.

Like, you know, probably gives you some idea on the different ways AI could impact the business. If you’re on the left side left side side of the spectrum, we wanna take cautious approach.

You know, the you could look at AI. The biggest impact of AI is really, you know, is improving member self-service, underwriting, and fraud. Right? And there are very well matured solutions who can do that really well.

And it could be about you know, if you’ll to look at members of service, which is where ours ex expertise is, you know, is often putting AI to on on in front of your contact center so it can answer a lot of calls, which is answering calls is extremely operationally challenging process. Right? Like, getting the frontline reps hired and trained and making sure the you’re managing the attrition and backfills In in a constantly moving world of knowledge where introducing new products with a new change going on and retraining is very, very, complicated process operationally, to make sure your members are always here getting the latest and greatest.

And, you you know, by the time you achieve that, you lose track of what could have been a higher potential, a higher value conversation. Right? Because it’s so operationally intense to drive this to even get the members the right answers to the question. But you forget the opportunity, you know, hey.

Those interactions, you could have probably outsourced some product or, you know, provided some detailed insight to the members that could help with the financial goals. That’s where AI could really make a difference. It frees you up from all this operation overhead that you can start really elevating the roles of the frontline staff to be more strategic and and and for the member and for the organization. Right?

So those are kind of the low hanging fruit, right, including, you know especially even there, there are so many phases. Right? Like, you know, you might have started the journey. I I thought thirty percent of you have started the journey there, but you maybe still not be agentic.

Right? Like, you know, continue to take that next step towards from generative to agentic is is kind of kind of low hanging fruit, including the employee AI, which is where you can have, you know, significant opportunity to make sure all employee knowledge is easily available for your employees, either frontline staff or beyond with the such a fragmented, you know, different places you have the knowledge. And then the right right end of the spectrum, which is what we are working with AltaOne and Phil, is they’re kind of taking quite a well, they’ve done all of this a few years back. Now

they’re really leapfrogging as fast as AI’s running. Right? So which is really kind of reinventing quite a quite a number of things, which is they’re putting a a bank GPT that can fully replace their traditional online mall banking. But, really, the the the the opportunity there is massive.

They are now with the AI enabling every interaction.

Able to seamlessly, very soon upsell cross sell products with every digital interaction, help members through financial journeys and financial goal setting, and and that’s gonna be pretty massive. Right? So, in in enabling all digital interactions to drive value in every single interaction. Right?

So yeah. So I I think it depends on which end of the spectrum you are. You could literally with with the bank GPT that Phil and team was gonna start, you can literally put retail banking on autopilot. Right?

I do see a a possibility in, like, you know, you know, two to three years where, a CFO of a credit union coming in and saying, here is my budget, here is my risk, here is my appetite, and here is the goals I wanna achieve. And the AI kind of really automatically spins up new products and, you know, helps achieve their financial goals and the business goals in a fully agentic loop. Right? That’s kind of the a a new horizon that’s getting unlocked with AI becoming better.

So, yeah, it it it depends on where you are in the journey and what’s the DNA of the organization.

Is kind of massive impact possible.

Thank you, Shri. The crawl, walk, run concept, very critical in that and different for everybody. Right? So analyzing that internally is is extremely important as you consider your AI journey for sure.

Real quick, I’m gonna remind everybody, including Phil here. We’re for today, the the voice AI phone call automation is what stood up. It’s what we saw here in the demo just a second ago. If we could progress to the next slide. Just some some context there as I get into the next question. So stood up in twenty four, powered by AI, obviously, integrations like Scenitar and Zoom in place.

And here, anyone who’s interested, the phone number, feel free to to test her out. Alright. Like, see what’s like yourself. We’ll leave it up for a second here, and then I’m gonna get to Phil’s question.

Before you do, we’ve actually embedded an Easter egg here. So if you call and you ask, what is Alta’s voice goal? She’ll give you an interesting response.

I love that, Phil.

That is so cool.

Goes back to that question about setting things up and how easy to use easy to use. Alright. So the question. Phil, what were the problems you were trying to solve with Voice AI?

Well, I touched on that a little bit. When I first got here, we were having a lot of long wait times in the call center. Obviously, you’re trying to keep your performance indicators, your abandonment rate, your hold times down, and the things that that that affect those are talk time.

You can imagine that if you got an agent who’s talking with a member who’s dealing with a really serious financial problem, and they’ve gotta spend time talking to that person and getting helping them get through that. In the meantime, they see the queue growing. They see the talk time their talk time going up. The hold time’s waiting or growing. The abandonment rate happening. There’s a lot of pressure on that person.

So we felt that if there was a way we could provide members a way to do sort of more automated capabilities for things that they just need a simple balance inquiry transaction, where’s the branch, so on, We would be able to allow our mem our members to have a longer talk times, unpressured talk times with our contact center.

Also, we’re trying to keep costs down. How do you deal with these periodic spikes? You know, they happen from time to time, and how do you know when they’re gonna happen? How do you staff that to happen? Sometimes those happen during the lunch rush when you’re trying to get people out to lunch.

It really makes it challenging to staff for that and to try to have a a team in place to handle those high volumes is is really tough. You know, we consider, well, why don’t we bring some branch staff in during that time to help? Well, branch staff aren’t contact contact center people. That’s a different ex skill set. So I just didn’t feel like that would be a good solution for our members either.

The other thing that was kinda sitting in the back end is that we had a legacy bank by phone system, and this thing was old. It was past end of life. The company was, you know, hey. You’ve gotta get off of this thing. We we can’t support it anymore.

And we were trying to determine what direction do we wanna go with that. We only had, I don’t know, a hundred and eighty to two hundred users. How do you maintain a system that’s only got that small number of users?

And then lastly, once we closed, members had to use that old bank by phone system, and they could basic they could do basic transactions on it only. We wanted to give them more of a twenty four by seven self-service capability.

And that’s when, you know, Stephanie had said, Phil, we’ve gotta go look at some AI solutions when that started us down our our path of looking at at AI voice.

And thank you, Phil. So both not only the consumer experience, but the employee experience being thought through. Very critical.

Sri, your your perspective on this. What what problems are FIs trying to solve?

I think, you know, lot of the kind of the problems that Phil mentioned are usually are are the first priority. Right?

But after you solve them, it just opens up a a quite a bit of an opportunity to do many other great things because you’re really putting a foundation solving that problem. Right? So which is your immediate needs, like, you know, call items are high, hard to train people, operational overhead.

These are all your immediate impact. And as you solve them, you also laid out a really strong foundation for many, many possibilities. Right? So and some of them, from my perspective, are, one, you laid out a foundation that could help prevent fraud in a significant way from voice. Right? So because now these are voice conversations.

The AI is involved in in deeply in how you speak and, you know, your your voice. And there is so much fingerprint data that could be used to, detect fraud even better.

You know, the the the whole concept of device biometrics, you’re balancing out, the friction and the fraud controls, you know, that that helps gives you a finer control to do it. It’s not like, you know, traditional, like, you know, members or, the IVRs that Phil talking about. The moment you want to do use phone banking, you say, oh, enter your member number.

I I called to ask you a loan rate, but, like, okay. Great. I’m gonna find my member number and call you two days later. Right?

You lost an opportunity, but AI knows when you ask a question where where is authentication needed, what is the risk of the question, how far the authentication has to go given the risk, and make appropriate choices to keep the member experience very, very good while protecting for fraud. That that intelligent decisioning is really hard in a static systems or old old old school systems, and AI can do that really well. So that’s the foundation for fraud. And then we’re also building a foundation that could help with, you know, enhancing this conversation to next level.

Like, you know, yeah, member call about a certain thing, you know, but AI could identify what they may not be thinking that is important to them and bring that up in the conversation. Right? So that that that can never happen in the old school system. Right?

So, you know, it’s it’s the the role of being a credit and a bank is being proactive, not being reactive. Right? You’re laying the foundation to be proactive in that relationship. Right?

So that’s the massive foundation you’re laying out that gives you more value beyond solving the immediate problem. So that’s that’s how I look at it. It’s a it’s a foundation for many things to happen. Once you train the your members through that, then you can open other channels and and whatnot.

Right? So which is you could have an agentic online, more banking, and all of that.

You know, you would have trained the members by then. Right? So yeah.

Thank you, Sri.

So, Phil, right back at you with with a big one. Right? Why why interface AI? What made you guys choose that?

Sure. And I can tell you that when we first started looking at AI, the drivers then are very different than what your drivers are facing today.

But back then, we were trying to put in automation to allow members to self serve. And anytime you push a digital solution to your members, you begin to minimize the person to person interactions.

And that puts your ability to grow the relationship with that member at risk. You don’t have that opportunity to offer them the next product. I don’t like to use the word cross sell or upsell. It’s got kind of a sell has a negative connotation.

Like to bring products and services to our members at their time of need, the time that it best fits them.

And you kinda have to be talking with them and understand that to know what is the next best thing to offer them.

And so that was the thing that we wanted to see in our in our AI solution. When we put out our RFP for our vendors, we said we need you to demo your ability to cross cross sell the bad word. Cross sell or upsell our members contextually, not just you’ve got an eighteen year old on the phone and the next product we have is an IRA. We want you to be talking with the member, and they’re asking you about a personal loan rate. And you tell them, by the way, did you know you’re preapproved for personal loan? Would you like me to help you get that now?

It’s it’s the it’s the moment of opportunity to actually improve the relationship there. So we added that in, and and as we went through and looked at the the vendors, Interface AI was really the only one that was able to actually do that contextual cross sell, upsell for for our members in that. And that was sort of a deciding capability for us on them.

The the understanding and delivery concept is is critical, especially in those moments. Right. Amazing. Sri, your take on, why financial institutions choose Interface AI?

I think, you know, we are by far the most ambitious in thinking how AI could reinvent banking.

You know?

So what we are working on today is not even in the road map of many others, but what we are thinking is not something others are even thought about. Right? So I I think we’re extremely visionary in what we want to do and where we can take this technology to really make a huge impact for the industry. Right?

So think that that that’s what people people like about us. Right? Like, even when Phil chose us, we were not doing upsell cross sell then, but we do it now. And he saw, the difference that we are thinking a couple of steps ahead always.

Right? So, I I think that’s what, you know, help helps us really stand out and because I truly believe in the meaning of partnership. The the meaning of the partnership is that you expect us to think about your business lot more proactively, how we can drive that, your member member satisfaction and your business, metrics. And and you want us to already solve those problems before you think about them, using AI.

That’s that’s what you’re partnering for, and I think we do that part really well. Right? So, it’s it’s probably there’s many other things and but that’s probably the single most important thing I kinda feel why, you know, partners like Phil and Altamonte have been choosing us for for for for years. Yeah.

A great tieback to that comment Phil made initially about thinking like a fintech or, like, taking that job that step.

Yeah.

Bill, could you walk us through the implementation journey? Sure.

And and I think it’s kinda funny that you asked that question because the rate of change with AI our implementation journey is not gonna be anything like anybody who’s just starting now is gonna have. But I’ll kind of walk you through it.

We started back with the little robot at the very beginning, where it was really, you know, just to use the term, a slightly smarter IVR. We had to try to determine what are all of the intents that our members are gonna ask and try to figure out how do we shoehorn all of these intents into a scripted IVR playback for those.

It’s the first time we’ve done it. We didn’t really know exactly what we were doing, and it it was rough. It was it was rough. That rollout was on the that was on the painful side.

But when we moved to v six, the rollout for that was way different.

We did all the testing that we could do in in on v six, and when we put it into production, members, not then granted they’d been using the system, it just worked better for them now, which would probably be more in line with the experience anybody starting now would have is that the agentic AI is much more natural. It’s much smoother in its response. You can give it complex things. I need can you give me the the balance in my savings and checking and then transfer fifty from my checking and then make a loan payment?

And it knows how to do those things. You don’t have it’ll just go and tell you those things in the order that you ask them. So it really makes it much easier.

The other thing that we kind of was a challenge is that our our staff is worried, are you trying to replace us? Are you trying to, you know, get rid of us by putting in AI?

The answer is absolutely not.

The successful digital implementation is a is is an augmentation to the human capability. As I mentioned before, we’ve our contact center has to deal with some really super hard cases. I I mean, I don’t know how they can I can’t talk about it too much? I don’t wanna cry in front of everybody. But there there’s some real heart plucking stories that they have to deal with in that in that contact center under tremendous amounts of pressure. And if we can reduce the amount of load that they’re having to deal with on a day to day basis to allow them to be that financial urgent care adviser, that doctor of finance, to be able to treat the symptoms of the problem that our members are having, we feel that that is the ultimate solution to to a higher level of member service.

It also provides a lot higher job satisfaction. When you walk home at the end of the day, I think, wow. I really helped a lot of people today. I just didn’t crank through a bunch of transactions for them.

Another thing we were concerned about and weren’t really quite sure how to deal with is we knew we needed to put in automated systems to attract a younger demo a younger demographic. I’m sure most of the credit unions these days are all struggling with a similar challenge we are.

Most of our wealth sits with an older generation whose they they’re not fond of change. They wanna stick with what they’ve always done.

And yet we have to put these new systems in to attract the younger crowd. That’s sort of a dichotomy that we have to face.

Well, we we I feel like we sort of solved that maybe unintentionally. When we put it in, I told you our members were really angry. So we needed to remove every possible friction point to allow them to get to the contact center. A lot of times, you put in these automations, you build three or four or five friction points they have to get through before they can make it to a human. We just took it out. Hey. Did they say representative?

Ask, you know, which what do you wanna do so I can get you to the right representative? And if they say representative again, call center.

And if call center can deal with moving them to the loans or collections or wherever else they need to go, remove the friction points.

Also, our older members, they like to have the human interaction. We have a particular member that calls and talks about her parrot. On payday, she’ll call up to get her loan balance, and then she’ll spend five minutes talking to the agent, you know, about her parrot, Fred, who had a cough. She took him to the vet, and he had a parrot cold. And, okay. Can you tell me the balance in my account?

So our older members, they want to have that human interaction.

But what we found is the the the queue became the friction point for them. Sometimes they just wanted to get their balance. It’s payday. I wanna go out with my friends.

I hadn’t we’re heading to bingo. I wanna know how much money I can gather with tonight. What’s the balance in my account? And I don’t wanna wait in line to get it, so they try it.

Okay, Alta. What’s the balance in my account? Sure. I can help you with that. They they put in their their information.

Alta gives them their balance. Oh, can you transfer fifty dollars from savings to checking? Yep. Transfer fifty dollars from savings to checking.

You’re good to go. Anything else? Nope. That’s good. So they they they take it and run.

So we’ve our older members have started warming up to it when they just need the information quick.

The other thing I would suggest is in the old school IVRs, you would give they would tell you, press one for account balance or say account balance.

People have been trained now when they’re talking to an automated system to just use the short version account balance.

While the AI can respond to that, sometimes there’s a lot of disambiguation questions that come in to figure out which account balance are you asking about.

It’s better to just teach them talk to it like it was a person. Hey. I need to know the balance in my checking account. Well, AI knows exactly where you’re talking about. Now I can help you help you go right to that.

So Yeah.

Phil, there are a couple of great questions, that the participants have asked. So one of the questions, was asked, how many seats does AltaOne see you have today in the contact center, and how has that number changed with the partnership with Interface AI?

I believe we’ve got thirteen members in the contact center.

And I haven’t checked in recent I probably should have coming into this call. But that hasn’t changed dramatically. We’ve made a couple we’ve added a couple of responsibilities to the contact center that we’ve put some people in specifically to help handle that. It’s it’s maintained consistent models since the time we’ve implemented.

Right. Thank you.

And another question that came up was Pisi, just to add context.

I think, you know, Phil, you can can be honest here. Typically, your goal is not to reduce headcount. That’s not how you design these for, you know, that could, you know, when if if you continue to grow, so you’re not scaling the team, right, which is which is Yeah. Another benefit.

And then also, when the back when the attrition happen, you probably don’t have to backfill with the efficiency you gain too. Right? So Right. Yeah.

Typically, most of the community banks, guardians we work with, they never, you know, have the goal of reducing headcount. It’s more like efficiency and repurposing the talent for doing other things or spending more time with the members and things like that, which is what we’ve seen. As yeah. It takes care of a lot of repeat requests, then your call time goes higher because staff is spending time on the more complex things, a longer period of time.

Right?

Yeah. Exactly right. Our our contact staff our contact center staff are hugely important to the organization. We would never think about taking them, you know, reducing that number.

Yeah.

Okay. Thank you, Phil.

Okay.

Wonderful. I’m excited to maybe talk to Fred one day and learn more about his financial journey. Sure. Sri, what is the typical implementation journey for our customers?

Yeah. I think as Phil rightly said, it has been changing with AI advent advancing very fast.

You know, in agentic systems, literally, you know, you’re entering a nondeterministic world. As in it’s not like a, let’s say, online mobile banking where, you know, it it it has a window and it has couple of buttons, and those buttons are always stay there. They always are blue, and then you can go click them. It’s it’s deterministic.

The online banking as an example is very deterministic. Right? Like, you build them once. They look the same.

They do the same.

But AI is a learning system. Right? So which means it’s constantly adapting to, what your members are asking within the guardrails. Within the guardrails, it’s constantly adapting itself.

So which gives you infinite possibilities. So it’s hard to test, you know, moving forward. So what we have done is we have what we call kind of test studio that helps AI show what we call kind of a golden dataset. It AI already goes and test, like, a a thousand iterations and list all of the things that pass or fail pass or fail for you.

And then you can just go in and add a few more things that you care about testing, and it’ll run. And you click a play button, it’ll keep running those all the time. Right? So so that ensures what what the industry calls it evals.

That’s a new way of testing, which is you define how the system needs to work, and then, it it test itself automatically against that instead of you testing and, you know, acknowledging whether it works that way or not. Right? So we have built a system where you can define all the test cases. It’ll already have a thousand of them.

You click a play button. It’ll run through those scenarios and come back, report whether you pass or fail. That’s the future of testing because there’s infinite possibilities of how you test. Right?

So, you know but, also, many things have become simpler. As Phil said, you’re scripting the responses. Now you just point AI to your website or upload a bunch of documents. It’ll learn everything instantly.

Right? And and you can now ask any question from that knowledge. Right? So and and then with a with a computer use agent, which is the inter intelligent RPA module, you also point it to a a your browser or a in a desktop tool.

Like, let’s say, your credit card credit card admin portal or your core banking desktop portal, and then, you you know, it can learn from how people are using it and start taking those actions as well by its own. Right? So the what used to take a a lot of time, doing these manual things, now AI has ability to kind of figure it out itself from website data or seeing how people use these tools. Or even looking at the user guide of your core banking manual, it can learn how to use that tool and start operating on behalf of a user.

Right? So that’s it’s almost like training a human being. Right? It’s like a digital employee.

You know, how I think about the implementation is now. Right? So very, very radically different from the past.

Absolutely. Like, Phil’s analyst too as for agent employees now. Right? It’s just self learning and the complexity simplified.

Yeah.

Even yeah.

Maybe before we go on to this next question, I’ll just kinda tack on to the end of, Sheree’s comment here.

Adjunctic AI, it generates a tremendous amount of data. I mean, it’s keeping telemetry of every interaction it’s having with members. So if it does make a mistake, you you can go into that massive amount of data. You can find that interaction, and you can follow the thread to see what happened, and then you can build guardrails.

And then but then you have to look for other things. Well, our person that that does that thought, hey. I can use AI to do this, and it built what he calls an AI judge. And so he’s got an AI judge that is monitoring the telemetry of our agentic AI.

Every interaction that it is doing, you don’t have to do this. He does it because he’s a nerd.

But it makes his job easier. But it it it monitors every interaction, and it looks for those things that could potentially start to become issues.

Then it surfaces those to him proactively rather than reactively from a member report. Then he kinda thought, well, one of them is good, but it’s a it’s it’s AI too, and it might make a mistake. So I want to have a second one. And if both of them come up with the same conclusion, then then I know that’s something I need to look at as as a priority. But it’s amazing, the stuff that that we’re capable able to hear.

That really amplifies the impact a single person can have.

Yep. Absolutely.

What kind of impact have you seen, and what is the adoption?

Well, I kinda talked to that a little bit.

You can see this chart that’s up here now. This is actually a report that you get off of, the interface systems itself.

And, you can see the total calls that the audio that the agent takes, and we’re talking to AI agent, Alta, and then what she was able to fully automate. But that doesn’t mean that these all of these calls automation means completed without making to a contact center rep. It means that they completed the intent that the member was trying to complete, and it may have resulted in a in a delivery to the to the contact center.

So we’re we’re getting pretty high automation rates, with the call that she takes for the things that it is that she’s trying to do.

The one thing that we noticed is when we started, and I’ve mentioned this before, is our our bank by phone had very low usage. I think a hundred and eighty. I think I said a hundred and eighty seven. It might have been upwards of two hundred. But we’ve noticed a a fairly steady uptick in our after hours usage, And I think it’s some of the people that have been that have used it while it was in full production and who are now opting in to use it are finding the value of it and realize they can get information anytime they want. And so we’re seeing our after hours usage ticking up, which we find to be very exciting.

Adopt adoption rate keeps climbing even though right now it still sits behind the phone tree. You have to select it to talk to it. That’s gonna change as soon as we can get the numbers ported correctly.

But we’re seeing people pushing the button to talk to it more and more.

And to answer that question earlier, our our staffing is held steady.

No layoffs, which we would not do, and then no new hires for the volume that we’ve had to deal with, specifically related to call volume. As I said, we’ve we’ve assigned additional resources or or responsibilities at call center that we’ve staffed up for. Our objective is is to scale the cost call center based on that need for the financial urgent care.

It’s incredible.

Sri, your thoughts on, is that similar? What are the kind of impacts are we seeing across our customer base?

Yeah. It’s it’s pretty consistent as you go into agent AI. The automation uptick is, like, ten, fifteen percent in in terms of the calls that that it used to, you know, take care of it fully itself. Right? So it’s a it’s a it it has a big multiplier effect in terms of the value, you know, we’re seeing in across the board.

Yeah. So there there’s there’s a lot more, you know, I I think we’re running out of time or want to continue hearing Phil’s. We can switch to other questions for him.

For sure. Phil, what best practices would you share with financial institutions looking to adopt AI? Okay.

The the first thing I would say is that if you don’t think you’ve adopted AI, you just don’t know you’ve you’re using AI.

I almost guarantee you every credit union on the call is using AI at some form whether they know it or not.

That’s what we found. We found that while we were trying to push it, it had become it had become pervasive, and people were using it all over the organization. We just didn’t know.

Once we started to kinda catch wind of that, we started to be worried about data leakage. How is data gonna are they gonna be putting member data up in these public AI solutions, and we’re gonna start having leakage everywhere?

So we stood up a formal AI governance committee. We established some clear policies and procedures and standards.

We immediately locked down any unapproved AI tool, including because we are Copilot shop, including unlicensed Copilot. It doesn’t reside completely within our tenant.

And, immediately, we started getting, hey. I use this AI for this thing and AI for that thing, and our governance team was able to sit down. We can now review those things, understand the business reason, information security can do a review, and we can make sure that we’ve got guardrails around protecting data leakage before we put that in place. So AI governance is is really important. Lock everything down and make sure that you are deciding to open it up, specifically.

I would start with the objective. What is it you want to solve and then get the AI solution to come in and and and do that? And we’re piloting new AI solutions to do that all the time.

And then always, like I mentioned, ensure you’ve got proper guardrails in to protect against any hallucinations and particularly data leakage.

We wrote, policies, and if you don’t know, AI can help you. We use AI quite a bit to help us with our policies.

And I know, Sri mentioned this. We are getting close on time, so we’re gonna be mindful of that. We do wanna take some questions, I think, if they hear. So, Phil, we’re gonna just pile them on. What what is your road map for voice AI and AI in general?

Okay. Now this is the part that think is the most important, so I’m gonna skip through some of the earlier stuff. We’re you’ve heard me say we’re gonna put, in front of the contact center. I I was just reminded that we have it in front of the contact center for the first hour and last hour of the day, and that’s a lot of where these members are getting an opportunity to use that. We are pushing to get member better member experience by continually educating her and getting her to be able to do, you know, everything that we could possibly ask about the credit union, she needs to be able to answer.

Anything that that that, any transaction that they would be able to do in the contact center, she needs to be able to do. But our road map for that is far, far, far beyond that. We are looking to we’re looking to allow the the AI to become that digital personal consultant, a financial consultant. I’m gonna use real quick an example.

My daughter and son-in-law, newly married, they bought a house. They’re over leveraged on their credit cards. Dad, what do we do? And I have to sit down Excel spreadsheet, put all the put all the credit cards in with the interest rates and the payments and try to figure out, you know, which ones would be the best way to pay off.

And then do we wanna pay it off the fastest, which is gonna potentially the most expensive, or you wanna pay it off with the least amount of interest, might take longer, or hybrid model? That’s a lot of work to figure out. AI could figure that out in, a minute. So those are the kinds of things that we want our AI, Alta, to be able to do.

And I don’t have to give it my statements because it has access to that. It has access to all of my statements, not just the AltaOne statements. I can ask it, hey. Do a full financial analysis for me.

Come back and tell me how I get out of debt, how I begin a savings plan, and how I can retire at this particular time. And it will go through and do all of that analysis and come back with a full scale plan with options for me.

Now they Interface may have talked to some of you about their outbound collection agent. Well, we would think collection agent would be part of your digital assistant. It would tell me, hey. You got a loan payment coming up. Do want me to make it, or do you wanna make it?

In fact, I would just say, you know what? You make all my loan payments for me and just tell me if I’m running into trouble while I’m having money. And so the the AI is going, hey. You’re a week out from a loan payment.

You’re not gonna have money based on your current cash flow. These are some suggestions you can do to have enough, and help me get through those challenging those challenging times. It’s proactive. It is going out to the member to help financially consult with them to make sure that they’re getting their payments on time, they’re getting their savings goals, they’re monitoring their spend behavior, habits.

Just that ability to just go, somebody’s taking care of my finances, somebody who’s way smarter about it than I am.

Now here’s the other thing. Our members wanna talk about the bird. You know, the parrot’s important to them. They wanna consult with the AI about, hey. My bird is doing this. What can I do?

And I would imagine that many people already have an AI on their phone. You might have you might use Siri. You might use, Gemini. You might have ChatGPT installed on your phone or Claude.

Those are the things that you’re using on a day to day basis. That’s that’s primacy for you. We want Alta to be that replacement. If you want to, we want you to take your ChatGPT app out, put your Alta app in, and get all of the same capabilities that you got with your ChatGPT, except that it brings to you a real world, digital assistant, the ability to manage and maintain your finances while it’s helping you out with all the other things, plan my vacation, plan my road trip, how do I plant my my mom’s yard and, you know, a bare root ochotillo, my mom’s yard.

It should be able to help you with all of those things because AI is already doing that.

Now my one last thing I wanted to kinda point out is way back in the olden days, we used to consider our our competitors to be the local financial institutions that were in our markets.

Then it’s then we had sort of the digital transformation time where we’re like, oh, shoot. Now it’s, we’re competing with the digital experience that Amazon delivers. We better step up and start trying to be Amazon or Applesque. That was a big term, Applesque for a while.

Then it’s like, oh, shoot. The fintechs are starting to come after us. We gotta, you know, protect against fintechs. Well, I would argue that today, our our competition, and I feel like it it could be a a challenging position for us is is and I just lost the names.

Can you believe that? Anthropic. I’m hoping. Anthropic and Okay. Microsoft and OpenAI and Google and all of those things, they have these AI engines that are deployed on hundreds of millions of phones that people are using on a day to day basis.

If any of those guys figure out a way to become fintech ish, that could be a serious problem for a lot of us that we they would immediately pull away the, primacy that we see in some of our digital banking apps where our members are and I would use I would say that traditional digital banking blockbuster model, that your member has to go to it to do things, and it have they have to do the work to do it.

Where the AI model is the Netflix model where AI is doing the work and bringing it to the member.

So I don’t we Stephanie did not want Ulta one standing at the edge of the pier watching our boat sail over to the horizon. Back in twenty twenty two, she was buying tickets to get on.

So I that’s sort of, you know, I think it’s really important that we we sort of focus on this, and that’s why we’re sort of working with Interface here to to try to realize that vision.

And and, Phil, your vision really solves some biggest problems in financial services, which is, the engagement is everything. If we can’t offer a product to our members that drives everyday engagement, we’re gonna lose out on on the business and opportunity. Right? So with you combining a general purpose assistant and a banking assistant into one AltaGPT.

Right? So you are driving engagement, knowing the members, and offering the right products and service at the right time. Right? So that’s that’s it’s a massive amount of engagement you’re driving.

And and and also, you’re offering this tool to somebody that who do not want to is is for somebody who is not a first priority to bank with Credit Union or Community Banks, which is the younger members. Like, the millennials and Gen z, it’s hard to track them. You’re giving them a tool that they needed today, right, and also making sure you the tool drives engagement. That is gonna solve many, many industry problems because we have the biggest challenge we’re working on as industry is generational churn.

Now you’re building out a tool that can actually, you know, excite them to use and then also drive engagement with regards to business. Right? So Yep. It is massive.

Sure. And I would even add on to that that I would like our members to share our AltaGPT with all their friends and family. They’re not even if they’re not members, let them put their accounts into it and let our AltaGPT run financial analysis on them. And every cross sell, upsell opportunity that they come to is an Alta cross sell upsell. Hey. AltaOne has this product that if you open an account, it would be great. And suddenly, we’ve got an AI agent that is out there trying to recruit additional members to bring that to show them the value we can bring to them as well.

Yeah. It it’s not you know, you’re directly with with that kind of positioning, like, you know, because, you know, ChatGPT and Claude, the Anthropic chatbots, they cost you twenty dollars a month to do, know, basic things. Now you’re combining the universal, you know, assistant with the banking assistant. You’re offering for free because there are so many ways we could upsell products and make money through those engagement.

Now you’re offering now now you’re not just reacting to a competition, but you’re going offensive in in kind of making a case that members actually don’t need to install ChatGPT and cloud apps to begin with if they have AltaOneGPT, which is massive.

Know, you’re protecting the business and driving business forward and attracting the right kind of members instead of being reactive. Right? So that that’s this massive, vision, Phil.

Yeah. And the value that it brings to the member itself, that’s really what we’re about. When I again, when I first started with Stephanie, I think she said was, I could give two craps about our our asset size. I care about how many members we have and how many people we can bring solid financial products and services to and and improve their financial position. So I think this would really fit that that with her very well.

Wonderful. We just have a couple of minutes, and and, Michael, we should wrap it up. It’s there’s just so much to take in. Right?

So there’s just a lot going on with AI. Hopefully, our goal was to, you know, convey it’s it’s not just chatbots. There’s a lot going on. And I think it, you know, it’s important to think through how AI will change our businesses.

And our goal was to give you some inspiration, some ideas on how some of the visionaries like Ulta one and Phil are thinking about it.

You know, thank you for joining us, and, you know, feel free to, you know, request for any demo using the QR code and everything. But we’ll continue to do more sessions like this that help you get the inspiration from people like Phil. I’m really excited for the series to come. More webinars to come as well. But thank you, everyone. I appreciate it, and have a great day. Phil, thank you for joining us.