How University of Michigan CU Uses AI to Elevate Member Service Without Losing the Member-First Promise

Welcome all. Good morning. Good afternoon.

Thank you for taking the time to join us for this webinar. How University of Michigan Credit Union uses AI to elevate member service without losing the member first promise.

We have exciting list of speakers with us.

First up is Sherry Wu. She’s the CTO of UMCU. She has been with, the UMCU for eight years. Sherry, can you quickly introduce yourself?

Yeah. Sure. Thank you very much. Welcome, everybody.

Thank you, Interface AI for this opportunity, to share what the UMC user journey with everybody, with this, University of Michigan Credit Union, UMCU for last eight years.

I’m holding the CTO role here.

Interesting enough, actually, this is my first, financial industry job. I’m always in the technology field, work on different industries, and they’re fascinated about, how technology technology enable the members, in their services, in the financial industry. So working for Credit Union, has been greater journey and bring we have been, quite a lot, new things to the members, enhance their, experience.

So looking forward to share more about what we had done, in the AI, chatbot, phone banking, with everybody.

Yeah. Looking forward to that, Sherry. So next presenter is Srinivas. Srinivas, please introduce yourself.

Thank you, Jack, and and Sherry for, being here with us. I go by Sri. I’m the cofounder CEO of Interface AI.

Personally come from a a Credit Union family myself. My father’s a star and scaled Credit Union to a couple of billion dollars and, helped them through the technology strategy in the early college years of mine. And when went on to work at Microsoft, part of the AI, team there on bing dot com and and later started Interface AI bringing both together for the credit union and AI experience.

Great. So she is gonna talk to us about, evolution of AI and show us some demos. That’s looking forward to that.

Thanks, Jack. John Taranzetti here, vice president of customer success at Interface.

I’ve spent a number of years in the conversational AI space and then doing a number of technology implementations and transformations myself on the customer side in the past. In my job, I get to work with all of our amazing customers on a daily basis and help them achieve their business goals. So I’m really excited to be here with all of you today and speaking with Sherry about, how we’ve been able to help University of Michigan Credit Union on their journey.

Thank you, John. And my name is Jack Chawla. I’ll be hosting this webinar. VP of marketing at Interface dot ai, And, I spent most of my career in communication and contact center technology space.

Okay. So let’s get started. So here’s the agenda for the webinar. So we’ll talk about Interface dot ai for, the people who are not familiar.

We’ll give you a good idea of what Interface dot ai does.

And then, John will talk to Sherry and talk about her experience with AI in general and, AI, from interface dot ai, how they’re using it, what is the impact, how the members are liking it, and and so on. But before we actually talk to Sherry, we, Sherry will show the demo of the two solutions they’re using, VoiceAI and ChatAI. So you have good grounding of what the solution can do, and then you can freely understand what the impact has been, for of the solutions for US Video Michigan Credit Union.

And then we’ll, after this, discussion, we’ll talk about AI in general, where it is going, how it is evolving, and how we are leveraging the evolution of AI to build very compelling solutions, for credit unions and community banks. And you’ll be excited about the things we are doing, which you might not even have thought about. So I’m really looking forward to, to this, discussion, and we’ll show some demos around what is coming down the line as well. And finally, we’ll go to q and a.

So please feel free to ask questions as we’re going along. If you’re not familiar with Zoom interface, at the bottom of your Zoom app, you’ll see an icon called q and a. You can click on it and ask, start asking questions. And we’ll answer them, either at the end of the, webinar, or we will directly answer it as we kind of go along.

Okay. So let’s dig in. So, actually, before we dig in, let’s take a small break and do a quick quiz to get a sense of, how, you guys are using AI right now. I wanna see get a feel for, what is your experience with AI.

And, at the bottom of the question, bottom of this quiz is a question, would you like to see a customized demo?

So if you, are interested, we can actually create a custom demo for you, which will actually work with your website similar to what the the demo will show you with the UMCU. So it might be a compelling experience for you guys to actually get a look on how AI can really transform, how your members interact with you on your website, mobile apps, and on the phone.

Okay. So, let’s see.

Please take a minute to answer the questions.

So fair amount of, about thirty one percent of you are using voice AI in your contact center.

About forty four percent, of the respondent are using chat AI for the website and our mobile app.

And, amazingly, very few people are using AI for employee augmentation.

Again, very few people are using AI for landing, and, quite a few people are not using AI at all. Okay. So I think in this webinar, you will get a lot of insights on how, UMC is using AI and, in general, what is the impact of the AI.

And looks like AI is a hot topic. Everybody’s trying to do something with that in the coming years. So I would strongly recommend, you engage with us and take a look at our products and demos so you can kind of educate yourself.

And it looks like their biggest pinpoint is, reducing the call center queue wait time and abandonment rate, improve the digital experience, which is, kind of a trend in the industry as, younger generation is, getting the transfer of wealth. They want to interact with the with the banks digitally.

And, actually, quite a few people want to figure out how to improve the productivity of the employees. Seventy seven percent.

Unfortunately, we’re not gonna show you this demo of EmployAI, but we will talk about what EmployAI can do for the employee productivity.

Okay. So let’s get started.

So Indevint dot ai has been kind of serving this industry for ten years.

Srinivas was one of the founder with Bruce, worked in India on, credit union equivalent. In India, we call them co ops, And he built the technology for, this company which he founded with his father. And he bought this technology to US to apply to the credit unions and community banks. And, we’ll talk about all the solutions we have built for the credit unions. And, about hundred credit unions have already deployed our solution.

I would say that our solution is most deployed in the industry.

The industry itself is very is in the very early stages of using AI, and we have the largest market share in terms of people using our technology productively every day in and day out. In fact, we have about one point five million interaction on our platform every day. And we handle about one point five billion interaction overall since we have been, since we started this company.

We have been mentioned by Gartner. We are, we are allied with the Credit Union National Association.

We are the sole provider of AI technology to a light solution, and we have won many awards, showing our innovative technology at Finovate and American Banker and various other conferences.

So now the question is the reason you are here, I think, is because you have interacted with the ChatGPD and seen the power of where the AI is going. Right? So ChatGPD is a amazing technology. I love it. I use it every day. And, the numbers are fascinating as well. So ChatGPD, they got hundred million users in three months.

And the latest data is about seven hundred million people are using ChatGPT every week.

And they’re projected to double it, in the next twelve months. So what that means for our industry, both for our employees and both and the members, is everybody wants a chat GBT kind of experience from the financials institution.

So it is a kind of becoming a race.

The financial institution who really provide the chat GBT kind of experience will win this.

It’s a it is becoming a must have technology.

And ChatGPD is kind of not only revolutionizing, the, the consumer experience with search and so on. It is some it’s a technology which will be highly adopted in the industry, in the in the, in the financial institutions.

Kind of similar to the impact of Internet and mobile phone, it will have a huge impact on our industry as well.

And Accenture did a study, did the analysis, and said ninety percent of the work in the credit union in the banking industry will be impacted by, AI.

Fifty four percent, which is more than half of the work will be actually automated and about twenty four percent and twelve percent. Twelve percent highly augmented by AI. Twenty four percent lightly augmented by AI.

So the impact of the AI is going to be substantial. It’s gonna be huge.

And I think it’d be used as a industry to really take advantage of what AI has to offer. And, we’ll talk about how we are leveraging this trend to make this, fairly straightforward for our industry to adopt AI in various different use cases.

So our as I said, ChatGPD is going to revolutionize the industry, and we are on a journey to give you basically a bank GPD. Bank GPD for the members, bank GPD for employees.

What that means is whenever the members interact with you, they get highly intelligent and intelligence at the level of your senior agents to interact with the members and answer the questions immediately. Right? So from a member perspective, if you can get answers to very common questions, even simple complex questions immediately through AI, which is human like, it is a amazing experience. And at the same time, if you if needed, you can actually talk to the human for very complex issues like, financial planning and so on and so forth. So combination of, highly intelligent AI and a human being will be the way you will kind of really delight your members.

And then on the employee side, employees, especially the younger ones, are used to check GPD. They will come to work, and they expect the similar kind of experiences. And that’s exactly what we are trying to do on the employee side as well. Basically, give them a very powerful, experience where they can, get answers to pretty much all the questions. Either it is coming from the system of record or your knowledge management, kind of combining all this and giving precise answer, grounded into your knowledge to your employees.

How do I open the trust?

How do I lock the card?

What is the, trend in the in my new membership? All these questions can be answered immediately through the employee experience, which we are building out, for, our industry.

So, basically, if you kind of think about it in a classic way, system of record is where all your, kind of information is. And then you have a lot of different applications and a lot of, you’re trying to fit intelligence in there. And then you have a lot of tribal knowledge where you’d ask Mary in the back office who’s answering the question. So we’re trying to bring all this together and provide a compelling experience at the fingertip of both your employees and members.

So what it means from the solution perspective is, over the ten years, we have built out what we call as a bank GPD platform.

It has the intelligence of one point five billion interactions we already done. It has the intelligence of all the, advancement which is happening in the AI industry. It is all kind of, bought together. And on top of that, we have built the security layer. So by the voice and device biometric, caller ID, ID foresight.

And this platform, is integrated into all the different calls, all the different iOS, all the different, online, banking, and all the typical system we use. So we are pre integrated pretty much with all the different solutions out there. And on top of that, we have built six solutions.

The first one is voice AI, which, UMC is is using, and Shri will actually show us a demo of that. And, keep an eye of how the, the voice is, AI is acting, and compare it to a human being and see how good it is. Right?

Similarly, Chat AI, which is similar to Voice AI, has amazing experiences. And, again, she will show you a live demo of, Chat AI on UMC’s UMCU site. And think of, when it is under when you’re seeing the demo, look at the questions which is the chat AI is able to answer, and you’ll be amazed on how advanced the, chat AI has become.

Then employee AI is, as I mentioned, kind of a, a friend of your employees. We will not be showing the demo of it today, but we’ll be more than happy to, kind of do a custom demo for you guys for the employee AI. And then fraud prevention AI, we talked about that. It is leveraging the, biometric and caller ID forensics and building the and bringing the intelligence on what how to counter the the caller, on, authentication depending upon the complexity or depending upon the risk level of the interactions, they are trying to do. So you don’t necessarily unless you want to challenge the caller if all they’re all they’re asking for is a routing number. And at the other extreme, if they want to do wire transfer, you wanna do, kind of double, triple authentication. So our fraud prevention AI automatically adjusts based on the risk level of the interactions.

And then we are working, in the alpha stage of this OLB GPT, which is, reimagining how the online banking can be done. And, it’s best to see the demo, and Sri will show the demo of this to really grok what, the OLB GPT can do and how amazing it is compared to the traditional online banking experience.

And then with the AI operator, we are reimagining the interaction between human and AI. Right now, generally, AI will transfer the call to the human, and, the human has to take over and complete a call. What we are envisioning is that instead of transferring the whole call, AI can hold the conversation and kind of have the human handle a particular, task.

And AI can continue with the conversation. And, again, when Sri shows you the demo, it will become clear on what this solution can do for us.

Okay. So with that, briefly, how the solution kind of sits into your, current stack. So think of this as one AI brain, which is sitting, in your in the in at the front of all your incoming member experiences.

On the phone, people can obviously call your phone number or text you.

On the mobile app or on the website, they can chat with you. You know, on the mobile app, they can be in the in message, and they can message you. And all this can be handled by AI at the level of, your senior, customer service member service agent, which means that it can easily automate sixty to eighty percent of the interactions coming in, at the quality of the level of the senior, member service representative.

So it means that it’s amazing experience. Obviously, humans like to talk to human, but nowadays, humans won’t answer from, immediate answers. It is, unfortunately, impossible to have that many human to immediately pick up the call. So AI can really, handle the kind of the routine request.

And then if needed, the call can be escalated to the, to the phone or chat system for the phone agents or chat agents to continue the conversation with full context transferred over. And there’s one brain is also helping your employees with all the different questions, and it integrates with your core banking, online banking, and, knowledge management system and LOS and all these different system to get personalized answer for the for the members and personalized response to the employees to help the the members when they’re interacting with them, be it in the call center, be it in the branch, or be it in an internal internal research or marketing campaigns and so on.

Okay. So I hope, that gives you a good idea of what Interface does, what our vision is, what our solution is, how we kind of integrate with your solution. So let’s kind of take a quick look at two of our solution, VoiceAI and ChatAI.

Thank you, Jack. Appreciate it. So, there is, so much we are doing. It was good to see that you go back and talk about all the solutions, and, it’s very comprehensive offering. It’s time to see some of these solutions in action. Right? It’s always a fun part.

You know, I’m gonna start off, showing you voice AI.

Pretty much, you know, you empty you, Sherry’s credit union here uses VoiceAI and ChatAI.

Wanna, briefly show you the products so that you kind of understand, what they do, and then it’ll be easy to follow along the conversation with Sherry and Jack later on.

VoiceAI, what it does, it’s pretty straightforward. It’s think about it as an AI system that picks up hundred percent of incoming calls and, has ability to automate forty, fifty, sixty percent of the calls depending on your member’s, you know, ability to interact with AI and willingness to interact with AI rather.

It has a ability to automate fifty, sixty percent calls and hand off the rest to back into your agent with a really warm handoff on context.

And during the process of automating, it can be in front of the pinned question, behind the pinned questions. And AI has a ability to even detect fraud, you know, and and prevent fraud, through very comprehensive, ways to authenticate the member, verify the member, and things like that. So that’s what voice AI does. We’re gonna show you a demo here real quick. There’s about four different scenarios, kind of highlight different capabilities of this AI. It’s pretty sophisticated.

You know, we start off, checking if I can afford a new iPhone that’s coming up on, next week. Right? And, and then, you know, see if, if I have enough balance and based on my expenditure, can I still go and make such a large purchase and whatnot? So I’m gonna have a conversation. You see AI guiding me through. And then I’m gonna try to, transfer some money, and also explore possibilities of, hitting my, savings goal, for a used car loan.

You’ll you’ll be seeing very powerful capabilities of what if scenario modeling and things like that. And we’ll also see third scenario where, I would explore to try to get a auto loan for a used car, and, AI is able to help me instantly find information from the website and guide me through the auto loan process. Then finally, we’re gonna talk about, a potential fraud scenario where you will see AI in action, which is kind of AI is gonna immediately recognize the need of, me as a member calling in to talk to an agent, without me asking for an agent. It is extremely sensitive, to member sentiment, to be able to balance efficiency and personal care for your members.

That’s which is so crucial, as, we’re creating and some community banks and try to balance the efficiency and personal care, which is often conflicting, but you need technology that you choose that understands the conflict and deals within a right way. Right? So we’ll see that in action as well. So I’ll leave this slide open here, and I’m gonna call the number here real quick and, put it on, loudspeaker.

And, Jack, if you don’t mind, when you can hear the AI in action, purely, just give me a thumbs up. Yeah? Sure. Thank you.

You can see the slide with the scenario still. Right?

Yes.

Okay. Wonderful. Let me call.

Hi. I’m Jessica, a digital assistant from Navy Federal Credit Union. I can help with banking services and general questions.

This call may be monitored for quality assurance. With the best experience, avoid using speaker mode. How can I help?

Hey, Jessica. I’m at Cypress, Texas. I’m looking to buy an iPhone. It’s cost me about eight hundred and fifty dollars. Do you think I can afford one?

To check if you can afford that iPhone, which costs eight hundred fifty dollars, I’ll need to take a look at your accounts.

First, could you please provide your member number? It should be five to ten digits.

Yeah. My member number is one six zero two three, and my SSN is one zero eight one.

You’ve got a thousand dollars in your vacation savings and seven hundred fifty dollars in your free checking.

So you have enough in either to cover that eight hundred fifty dollar iPhone. Which account would you like to use for the purchase, or do you have any other questions about your accounts?

Yeah. You you know what? I would rather go, get cash from another account. Can you tell me, nearest ATM or a branch, close by, here?

The Cypress Texas branch is located at fourteen thousand one hundred and forty Cypress Rose Hill Road, Cypress, Texas seventy seven thousand four hundred and twenty nine. It offers a twenty four hour ATM and Navy Federal ATM.

The lobby hours are Monday through Friday, nine AM to five PM, and Saturday nine AM to two PM. Want me to text you the address and hours so you have them handy?

Standard text message charges may apply.

Not right now. Let me go and send, about, fifty three dollars from my, checking to savings account. Can you help me with that?

To proceed with sending fifty three dollars from your checking to savings, I need to verify it’s really you.

I see you have a couple of phone numbers and emails on file. Could you tell me which one you’d like me to send a onetime password to?

You can choose between the phone number ending in Let’s go and send to my phone.

I sent a one time password to your phone number ending in eight two eight three. To validate, could you please enter the code you just received?

The code is one two three four five six.

The transfer of fifty three dollars from your free checking to your vacation savings is all set. Now you’ve got a total of one thousand fifty three dollars in your vacation savings and six hundred ninety seven dollars in your free checking.

With that iPhone purchase in mind, you’re still looking good to cover the eight hundred fifty dollars.

Anything else I can help you with today?

No. I’m thinking about buying a used car. It cost me about ten thousand dollars in a year from now. If I have to save, ten thousand dollars in my savings, let’s say, in a year from now, how much should I be transferring every week?

To save ten thousand dollars in a year for that used car, you’d need to set aside about one hundred ninety two dollars and thirty one cents every week. That’s calculated by dividing ten thousand dollars by fifty two weeks in a year.

I Want me to help you set up automatic weekly transfers to make saving easier and less prone to being forgotten?

Yes. Let’s go and set up that automated payment.

You’ve now got a plan in place to save for that used car. I’ve set up a weekly transfer of one hundred ninety two dollars and thirty one cents from your checking to your savings starting tomorrow.

Your savings will have one thousand two hundred forty five dollars and thirty one cents, and your checking will have five hundred four dollars and sixty nine cents.

Anything else you’d like help with today, or are you all set?

You know what? I’d also apparently wanna explore a auto loan for that used car. Just wanna check the rates.

What is your current rate, for a ten thousand dollar auto loan and new loan?

You’re looking for auto loan rates for that used car you’re planning to buy. Right?

We have competitive rates for used car loans. For a ten thousand dollar used car loan, our current rate is around five point five percent APR for a sixty month term. However, rates can vary based on your credit score, loan term, and other factors.

Would you like me to give you a more precise quote or explore other loan options that might fit your needs?

Yeah. Before I probably proceed on applying for this, can you tell me what documents, I need to have ready, before I proceed probably applying on a branch or online?

To apply for that auto loan for your used car, you’ll typically need a few documents ready.

These include proof of income, like your recent pay stubs or last year’s tax returns, proof of where you live, such as a utility bill or your lease agreement, and some form of identification, like your driver’s license or state ID.

Since you’re looking at a used car, we’ll also need some details about the vehicle itself, including the van, the make and model, and how many miles it has. I can send you a quick summary of all these documents via text if that helps, and standard text message charges may apply. Would you like that?

Not right now. You know, can you tell me what are the top three most recent transactions on my checking account?

The top three most recent transactions on your free checking are a debit of nineteen thousand two hundred thirty one dollars on August twenty eighth, described as transfer to account five hundred and sixty thousand one hundred and twelve, a debit of five thousand three hundred dollars on August twenty eighth, described as transfer to account five hundred and sixty thousand one hundred and twelve, and a credit of one thousand dollars on August eighth, described as Social Security deposit for July. Want me to explain any of these or check your current balance?

Hey. Some of those transactions you, you shared that you shared just now, I didn’t do them. I suspect there’s probably a fraud. Can you help me, with that right away? I’m really worried.

I’m so sorry to hear that you’re concerned about potential fraud on your account. I’m going to connect you with one of our fraud services representatives right away. Please stay on the line, and they’ll be with you shortly to help resolve this issue and ensure your account is secure.

Great.

So hopefully, you saw the AI in action there.

You know, for the sake of the demo, some of that intentionally slowed it down for everyone to follow.

Hopefully, you saw AI, was able to kind of help me with, see if I can afford, to buy an iPhone, you know, based on my expense patterns and things like that. I was able to also remember my location. Now when I said in the beginning of the call, I was at, Cypress, Texas, it was able to then when I asked for an ATM or a branch, it remembered that location offered me something close by. And if you noticed, it was pretty intelligent in kind of the follow-up questions, telling, hey. Can I send you this text message with all the information and whatnot?

And then when I went into transferring money, hope you notice it was able to guide me through my savings goal. And, when I was asking for auto loan, it had the ability to instantly look up the knowledge on the website and respond to a lot of questions about, you know, preparing to bring the right documents or, is there a, is there an auto loan that makes sense for me in terms of rates and things like that was able to guide me right away. And finally, when you, when I asked about some, potential fraud, on my account, it was, it was very seamless to be able to hand off the call to an agent, without me even asking.

Right? So, that’s kind of the, part of, the AI. This is kind of the next version of the AI that UMC is using. UMC is on version five.

What we just showed you is a version six. It’s a little more latest. I just wanna call that out. And, with that, I wanna show you, the chat AI demo.

Chat AI has multiple modules.

You know, there’s about four modules on Chat AI.

Smart discovery is a Chat AI that, answers in front of the pin question, and smart transaction is a Chat AI module that answers behind the pin question. And the smart conversion is a module that actually cobrowses with you along the journey.

Either you’re applying for an online application or shopping for a new product, it actually cobrowses with you. Gone are the days there is a human need to cobrows. AI can do that itself. We invented that technology. And then there’s finally upsell cross sell. There’s an AI that actually drives product recommendation, financial insights, and organization.

UMC uses these two products today, in front of the pin, behind the pin. They’re only live on the in front of the pin, and we’ll show you that in, action in live. So I’m gonna request my colleague here, Laura, to go and share the screen. So we’re gonna, hop into their website here real quick, and, you can try this yourself as well firsthand.

There’s nothing like a firsthand experience. So, if you see the right bottom corner, you can see that, AI in action. So you can pop it open, and we’re gonna start asking a few questions and show you how powerful it is. So, we’re gonna try asking about auto loan rates similar to what we showed you on the, call.

What are the auto loan rates?

And, what documents should I need to bring to apply? So, immediately, here, the AI was able to respond with the actual auto loan rates they offer, what documents I need to bring in. And, if I come below, there is multiple sources. It shows at different places it actually looked to find this information summarized for you.

So you don’t have to do it. Look at that. They had to go through three three different links to find that information. Imagine trying to do that, as a as a member.

It’s challenging. It’s just not EMCU. It’s just the way our website is structured is not, you know, easy to find information often.

So, you know, a, I was able to go through all of that as a true assistant and find the details for you. And then when you, come below and click on that blue link, it automatically navigates, to that page where the information was, taken from, automatically scrolls, highlights in yellow, the the rates and everything. Right? So the AI is automatically navigating on behalf of you, to to, show show you exactly where you found the information.

So we’re gonna try a couple of other things.

So, you know, we’re gonna try to do, some, math here similar to what I did in the call. We’re gonna say, hey. For a twenty five thousand dollar car loan, what would be by monthly payment, based on, your rates? Right?

So comes back, tells me, you know, you don’t need a a calculator. You can just talk to AI and figure things out. Comes back, tells you, hey. Your monthly payment would be four hundred dollars.

Right? So now, of course, it also offers you a link if you wanna, you know, play with it on a calculator. Right? So, it’s very powerful.

I I strongly encourage you to go to u m c u dot org and try, hopefully, Sherry doesn’t hit me, to drive a lot of traffic to the, to their website.

But, we’ll try, one more question here.

You know, you MCU savings account for students.

Do you have anything specific for students? Imagine trying to do this on the website often. You gotta navigate the whole site to figure out where, the student offers are. But here, you know, just a couple of seconds, AI was able to find that information for you and and, bring that up.

So the beauty of this AI that we’re showing you on the voice or on the chat is there’s zero training. Like, you’re not sitting and custom scripting any of the responses, which is the case with a lot of other AI, options out there. With us, you feed the knowledge and, integration, and it’ll it is smart enough, already trained enough. It’ll figure out autonomously how to complete your request.

It’s pretty powerful.

You’re not, setting a predefined workflows and kind of sit and train manually and things like those days are old and gone. Right? So, we’re gonna try to ask a follow-up question here real quick.

Based on your offering, how can I consolidate my student loan? Right? So, we’ll see. It gives you a little more details and help you with that, question as well.

Can consider debt consolidation loan offered by MCU. This allows, you know, here is how you can the steps you can take, how you manage that. In at the bottom, you have a blue link. You click on it.

The first one is it was already looked at multiple places for you. So you click on that, and it’ll automatically take you to that page and, you know, you know, highlight that information for you. Right? And then automatically scroll as well.

So hopefully you get an idea how powerful, this AI can be, and, now we’ll show you the behind the pin experience.

You know, and and EMC is in the process of enabling that. When they do, on the same chat, when you ask any behind the pin, it’ll authenticate right there and then and answer the questions too. Right?

And and, you know, for now, we’ll show you their demo setup.

Let’s, go ahead. We’ll show this on an interface bank. Don’t worry. We’re not in the banking business.

This is type of hypothetical bank.

If you see, there is a right bottom corner. There’s a chatbot already. So click and pop it open. And, we’re gonna ask behind the pin question. Now you’re already on online mall banking, so you don’t need to authenticate. So we’re gonna try a few simple things, and we’re gonna, tune up the complexity.

We’re gonna try four question. The final one is gonna be the most complex. You’ll never be able to achieve that online mobile banking.

Kind of shows you how AI could be a true companion. So the simple one to start with, we’re trying to transfer money. Right? So it shows you, hey.

Is this what you wanna do? You click a button. Yes. Let’s do it.

You’re done. Gone are the days you gotta figure out where to click. Navigate a form. Fill up all the details.

You know, you just type it in, or even you can speak into it. You see the mic button right bottom corner? You click on it, speak to it. It get takes care of it.

Right? So now let’s try something a little more complicated, to do in online mobile banking. So we’re gonna try to dispute a transaction, which is specific transaction, thousand dollar charge.

It’s painful to find that transaction in online mobile banking. Gotta navigate your statements and figure out and click on those. But here, the moment you asked, it found the transaction, said, hey. What would you like to do with this transaction? You wanna dispute, put the card on, freeze the card with a click of button again. You can complete that.

You know? You see, like, how, you know, AI can be a true companion there for online banking. And then we’re gonna try something a little more complicated now. Try to make a a credit card payment and, you know, scheduling a payment here.

It comes back says, you know what? I can do that, but I I I recommend you don’t do it right away because you have imbalance in your in a, you know, in your cash balance. So you may wanna wait it out. So because you have upcoming expenses, why don’t you schedule it for September three September third.

Right? So you have a button at the bottom.

You click on it, and you schedule it. So the AI was able to guide your members to take right decision providing financial insight. Now we’re gonna try something very complicated that’s almost impossible to do in online mobile banking. Kinda shows you how AI can be a true companion.

So, we’re gonna ask for, one of the popular questions we all have.

How many streaming subscriptions have I signed up? You know?

For the last one year, how much I’ve been paying? So imagine trying to do that on ML banking. It’s almost impossible.

So here we type it in, and AI comes back and says, hey. I can pull all this information and do the analysis. Can I proceed? You say yes, and it takes over control from you as a member, automatically navigates your online mobile banking, downloads each statement for each account, and analyzes those transactions and creates a beautiful summary, to show me right there on online mobile banking exactly how much I spent.

You’re still on online mobile banking. Right? It didn’t go away. So AI took the freedom to take full screen real estate at the right time because the information, was, more for a small screen.

It made that decision automatically and constructed this graph on the fly. Nothing is predetermined.

Right? So if you click on the right top corner, the arrow mark, it collapses, and you’ll be able to see the online mobile banking right there. It can go back into your AI full screen anytime.

So, you know, pretty much, this is the way we want to plug into existing investments, to make sure AI is working seamlessly. Right? And then in the future, as your members and, and and you you all get comfortable with AI, we are we already have a, online mobile banking system that’s fully AI first. You can replace that.

Right? So, that’s the kind of a little bit of a demo for you. Of course, I highlighted some of the components and UMC doesn’t have yet. They only use in front of the pin.

They’re launching some of this behind the pin as well. Also, on the voice AI, they’re on a one version previous to the latest. They’re gonna be live on the new version sooner too. Hopefully, that gives you some idea about their differences.

And, Jack, back to you.

Thank you, Sri. Wow. Mind blown. Right? This is playing GPD in action. Yes. I hope, you guys got a sense of what is possible.

Right? So if you cannot really think about it, it’s impossible to do all the things Sri showed you with the traditional WIMP interface, Windows, icons, menus pull down. So I think, a lot of smart people in the in the technology industry are saying, WIMP is going away. We totally agree with them, and I think the conversation will be the way to kind of interact with the financial institutions, be it calling on the phone or the more digital experiences.

And our technology can really power that conversational experience.

So with that, let’s talk to Sherry, and, learn how they are using how Sherry is thinking about AI and how they’re using Interface dot ai.

Okay. So let me hand it over to John, and he will interview Sherry, and, Sherry will share her insights.

Awesome. Well, thank you so much, Jack. As we mentioned in the beginning of the call, we’re very excited to have Sherry here with us from University of Michigan Credit Union.

University of Michigan Credit Union, one of the things I love about this in this in the statement here, you know, it started with twelve members and nine dollars in assets in nineteen fifty four. Today, over a billion dollars in assets and a hundred and twenty thousand members. So the work that Sherry and her team are doing are really, at a significant scale.

So, Sherry, as CTO, you’re obviously intimately involved with how AI can improve the operations of UMCU. You know, how is the business strategy at UMCU delivered more effectively using AI?

Yeah. Sure. So, for us, in, in the Explore AI, we explore in multiple areas.

Started with the enhanced member, experience, on the website or on the phone. So partner with the Interface AI, we launched the, chat AI, which is a a smart discovery and a voice AI, as well. So by having those, we’ll provide in the real time answers, to members anytime. Right now, they’re just doing a busy hours. So before we have this capability, members are calling us of ours, you know, the content center close, at six thirty, on the weekdays. And so they were now able to access us. Right?

So also by having the, the voice AI on the phone, reduce the wait time for the members, and then so wherefore you improve their satisfaction.

So that’s one area. Second area, we also, looking at, it’s a streamlined loan origination with AI. We have traditional loan, system, iOS system ourselves. So we’re looking at to enhance that whether by ourselves or purchase, third party, utilize AI, the, AI driven decision system. So make a quicker way, effectively, decision, for the membership.

Then we all know the for our team members. Right? We all utilize things like a chat TPE to compose emails and things like that. So we encourage the member team members to utilize those to make, things quicker, what they do.

So, improve quicker, right, decisions. Then it can free up time to do something else.

And, another area we’re also looking at it is, explore the AI driven, ticketing system. So we have a a ticketing system right now mainly used by IT, but also by other department as well.

It’s, it’s a quite old system now. So we’re looking at, one of the system. It’s not only have a newer capabilities, also have AI capability.

So with that, the mum team members, when they ask questions, AI can answer some if questions already be asked before, or, for an IT help desk. And by utilizing, like, the tool, they can get to the quick questions answered for them, and then they can pass on. Right? So that’s another example of AI.

So for similar to what you described earlier, employee AI, we also evaluating what we do regarding employees, knowledge base. So right now, knowledge base is all different places. Right? The procedures, some are here, some are there.

Take a time for team member to switch this right procedure I should follow. So we’re looking at the building, AI system where, you know, can you just like a chat to a member use, right? Say, oh, how do I do this? And then the answer come up right there, for them.

So saving time for them. So, so that’s five different areas, all whole, you know, serving, team members or membership to streamline the process and then make, their life easier and, improve the satisfaction.

Yeah. And what I love about this is the different dimensions that you mentioned. Right? You have both informational and transactional and AI for members and for your employees. So there’s really an ability to transform all components of the business using AI.

And in doing so, as as you see on the screen here, as Sri mentioned, Interface currently works, with UMCU with VoiceAI since twenty twenty four and ChatAI since twenty twenty three. And as part of that integrated, seamlessly with, their core and with other, elements of their ecosystem. So we have Scimitar, LTRP, and Cisco all as integrations that were enabled as part of our discovery and implementation process with UMCU.

One of the really cool things is, is Jack and True were alluding to earlier is hearing the technology live. So I don’t do it now. But after the, the webinar, certainly invite and encourage everyone to test out and give, Vic a call. I think Sherry and her team have done a really amazing job and have created a really, you know, top tier program using all of these AI assets. So give Vic a call, the numbers on the screen here, after the webinar, and see how great it is for yourself. So, Sherry, the next question here, what are the specific problems you were trying to solve with voice AI as a channel?

Yeah. So, for us, you know, before we have the Vic, when whenever we launch a new product, there will be a huge surge, from the member calling, to know about the product where they don’t know how to use in it. And so that causing a long wait time, and a higher bending rate due to we only have, you know, fixed number of agent to answer those calls. And so we were say either, you know, implement something. Otherwise, we, we, you know, hire people seasonally or we have a third party do the, call flow. So those are all, expensive and then, take time to implement.

So utilize the voice AI. We talk about, you know, can, can make that things much quicker.

Reduce the member, frustration from the wait time. Right? We said if you are doing a high call volume time, you know, if you have waiting for a long time, that is frustrated for the, members.

And also or they simply drop off the call. So we track that abandon rate from the contact center as well. One of the metrics we’re tracking.

And so with with the voice AI, we allow the members to ask in those simple phone questions easier just to chat. Now only transfer to the agent if the answer cannot be answered or they really want to talking to agent. And we made a flexible not to be say, oh, the voice AI will take over the whole thing. Instead of what we did is, like, it’s become one of the options in the country.

So it’s a press one to talk to Vic, and then you press two for regular member service. So we are not enforcing everybody to using VI, but give a try. Once you try the Vic, if you like it, continue. If you don’t like it, you can always say, talk to agent.

Get out. Right? So we want to make sure we serve all members to how they want to be served and not forcing them to do something. Right?

So that’s another example of then, agent, by, having freed up with the simple question. They are taking all the more complex questions like a fraud and things like that. Right? And then or upsell, to to be even really better, you know, member service, to have the time to doing so.

So, you know, talk about the limited, service hours. Right? This, opened up twenty four by seven. We recently launched the twenty four by seven, on the on the phone side.

And, you should be every time. I also oversee the our, the phone phone system. Every time we do a phone maintenance of ours, we will open up the contact center for testing the call. Every single time, we have a real member call you.

What that means, they do want to talk to you after hours, but you are closed before. Now we’re open. And this can serve to them certain degree. If not able to answer, they can call back.

But, hopefully, we’ll have more and more question be answered, so they can, you know, serve serve the members twenty four by seven.

Yep. And I love one of the things you mentioned in terms of thinking really strategically about how to place AI so the people are excited about it and they don’t view it as a blocker. You know, one of the things we hear from a lot of different CEOs is, well, I’m afraid of using AI because I know a lot of my members or a certain part of my member base just like talking to someone. But the way that, you know, you and your team have implemented it, it really allows the best of both worlds.

It gives people who wanna use the AI that option and doesn’t alienate the people who just wanna speak with someone. So there’s a lot of really creative and smart ways that you and the team have pulled this together that is allowing your AI transformation, but is also making sure that members who really want that personal connection with the CU can continue to get it. And as you thought about the partner selection process, how did you go about that process? From the moment you decided you really wanted to implement an AI solution, where did you look, and why did you choose Interface ultimately?

Okay. So, so for partner selection process, this is not just a space speak to AI. Right? Almost to all the vendor selection process.

I kinda follow this.

So, one is attending the industry, conferences.

So I’m not limited to myself to just credit union. Right? You know, financial industry. We’d know the bigger banks have, you know, we’re wasting two things.

So we want to make sure you’re attending those conferences and talk to colleague in your credit union or banking industry, to know what they’re using. And, linking is another source I’m following, you know, the vendors there to see what, new features they put in or new versions they put out. That’s example. So first, we actually I met, Sheena when, he presented during the COVID, CUNA Technology Council, a conference we had to force that conference be online.

I think that was first time the online conference versus in person, and, Srini, demoed the, the solution at the conference. I really like it at the time and reach out to the team right after the conference.

Sherry So Sherry, you had another AI chat on the website that we it does well. Yeah. You’re already even before pandemic, you’re already using AI for Yeah.

So a little bit of it.

Yeah. We are actually just talk about, like, some you mentioned, like, some of the maybe hesitating or afraid. We, we actually implement our first chat.

We I I would say seven years ago.

Right? So we had the data right away, and we see some limitations. And then once our contract is up, so we said, well, he likes to reevaluate to maybe found a different provider. So that’s how we you know, I’m looking to see what else are there.

And one of the criteria I like to have is the vendor serving multiple purpose. Now just solving one user case. So whether you are on the phone, website, and I’ll send inside online banking. Right?

So that way, the member have one unique, unify the experience as well as the team members. Right? You have logged in one system. They’re already familiar with serving multiple channels.

Right? So that’s another criteria I’m looking for when you value the vendor.

So by talking to, to the vendor, not only look at what they have demoed, make sure it’s all, you know, what is is they said is what they have done. And And then talking to other reference, if they have the reference, if you are not the first one, talking to other unions, they will give you an honest opinion about the vendor. And then how about the project implementation process? Right?

If not to create a product, if you don’t know get to how to get launch it. Right? How to get on time, launch it. So that’s not a big area talking to and then timeline.

And, integration is huge area too. What all the system this vendor already implemented?

You’d most likely, you don’t want to be the first one, but sometimes you don’t have a choice. But asking those questions, what are the integration you already have and then what you were planning to do? And the vendor road map, really critical. Not only is it what they have done, what they plan to do in the next year or two years.

You don’t want to pick a vendor, sit to steal. And when everybody else move down, this vendor just sits still. Right? So that’s another huge criteria, when you’re looking for, an any vendor.

Right? So not just not specific to AI, but AI is not even more important because how quickly AI involved with the technology changing.

So it’s even critical for the vendors to be staying in the front, on that. So why interface AI can talk about a little bit already. Right? So it’s, meet all those criterias.

It’s a combination. It’s not a science. It’s just, you know, and I also involve is out, when I initial selected, then I have, my senior leadership all joined the the, the, you know, demo with me together asking questions. And then, we, mentioned we already had a a a chat on a website before.

And then we said, oh, not only we want to replace that, we also want to have live chat.

So how about we do this together? Right? So the member not only got a a new, chatbot, they also able to do live chat. So we select the Ultra p at the same time, and we have that integration going on.

So we will have the AI chatbot. A member can transfer to live chat if they pick two or actually can transfer to text with the auto p to the to text. So exactly, sir, meet where member want to be served. And, so then, so we, you know, demonstrated the flexibilities.

Right? So I will talk about a little bit later. Certain things that we do is not typically interface AI do the order you do, how or how you do it. We have strong opinion how we want to do things, and we make that happen.

Right? So and, I have many of my team members from the project actually on the call today. So they definitely, played a greater deal of how we think it’s the best way for us, for our credit union, for our membership. Right?

So now just listen to the vendor say what this is supposed to do. You have your own opinion, your opinions to the final decision because it’s you’re the one serving to your members. Right? So and then talk about agenda AI.

Right? You used to be a a journal AI, and we’re providing the way we’re doing this, but then, with the new word, it’s a agentic AI. So, I know interface AI is seeing the roadmap doing that with the, the next release and the for this year, later this year, or next year. So that’s another example of showing that you look how what the vendor’s doing, to make that decision.

And and, Sherry, huge kudos to the team and your team and and having the clarity of what the members need and what your members need because no two credit unions are same. Right? So every credit union, AI gen is gonna be unique, just like their members are. And and, you know, and and huge kudos for your team to have the clarity in guiding us through to and and we being on the flexible on the other end. So just worked out well as a partnership.

So appreciate appreciate your team and you as well. Thank you.

Yeah.

Yeah. It’s absolutely both an art and a science. There’s best practices, but as Shree said, it totally comes down to what your members want, how you wanna serve them, and what you want that experience to be like. So as you were thinking about bringing that experience to life and you had to kind of make this all work and pull it all together, could you walk us through what the implementation journey looked like for you and your team?

Yeah. Sure. So first of all, we, talk about the project governance and, we, served what we did is synthetic, we have dedicated the project managers, from UMCU. And also I know interface AI also have a a project manager.

It’s so it serve as a single point of contact. So those two work really close together on the monitor, the issues, the timelines, and so I’m you know, I had the project manager, Carrie. She’s online with us today. And, no, you know, we are so, she had the great inputs on everything we have done.

And, project community, it’s not this AI project is definitely is not, a IT project. Right? Or single, department project. So it’s collaborated with multiple departments.

For us, it’s IT and, contact center e service. So the two two debate department, and I have the COO oversee the contact center e services, co lead of the project. And, so we are the cosponsor leading the project. So we make sure, not only technology wise is is mostly, but also really serving what the members are looking for.

Right? A weekly meetings, we have project weekly meetings to talk about the requirements, the testing, the timelines, and the quick resolutions.

So another, tool, no in, interface AI, use is a ticketing system. I really like that. So it’s a single point where system goes. Right?

Talk about where the issues and then the other resolutions, ticketing would receive a response right away saying, okay. Would it resolved or not? And then then in a weekly meeting, we talk about those high, important tickets. We call the showstoppers.

Are they really who showstoppers or they can’t live without, we can move on. Right? And, clear definition of scope. Right?

Good enough. Right? Certain things, I mentioned earlier. So it’s us, at the Credit Union, make sure in the end determine whether we’re ready to launch.

Right? It’s not a vendor. They they may test it. It’s, oh, it’s ready. But if we don’t feel comfortable launching, it’s okay.

Right? So that’s what you don’t like about. So one example I gave is initially interface that you guys have the voice, for the, with it’s a little bit robotic. Right?

And then you say, oh, actually, other client to launch with it. But we said, no. We don’t like it. We want a different voice.

So we said, okay. We wait.

Working with you guys to get a new voice.

Yeah. Right? So that’s example of something maybe simple, maybe small, but we have strong opinion on. Say, no.

We want a different voice. So I think that’s something really paid off. To know how we care about our members, we want to the best. We know from other experience we had how could could the picky our member is.

Right? So we want to give them the best experience from the day one. We don’t wanna say, oh, we launch something or we can improve it. Certain things, it’s okay.

It’s we can launch and improve later. It’s agile with the face approach, but certain things are not. And, talk about the the even the chatbot. We wrote the chatbot first versus use a voice first.

For traditional, interface AI is always go to voice because most of the credit union do not have, a, chat on the website. But we do. So we you know, our contract is up. We need to replace that first.

So we switch over. So, okay, we do not want to we want to using the, let’s put on the, smart discovery on the website first. So we did the switch to that order as well. Right?

And then we want to introduce Ultra p. Let’s say, how about let’s get here the replace the, the current chatbot right away so member can utilize. Then we add the Ultra p for the, integration, for the, you know, live chat and the live or texting services as well. And, so we need to work with the vendor.

Talk about the the the importance why you wanted to order the way you do and then get the buy in from the vendor, say, okay. Why this makes sense? Right? So face approach for voice AI.

Again, we don’t do a bigger bad say, just put on the voice. Instead of what we did is a face. What we did first one, we put on the off hours only. We said, let’s try it out.

Let’s remember try because we know they always try call us off hours. So let them try. And now before you should be god the rest. Oh, it’s we’re closed.

Sorry. Call us back tomorrow. You said that they got, oh, you know, then they can answer those questions of hours. Then we replace our, older technology IVR.

And then when IVR, we said that when they called IVR, they had those speed on their speed dialing. Right? Some of the old mem you know, membership. So we changed that number to using the, the Vic.

Right? So we did that second. And the third is when we feel comfortable with all those two, then we said let’s put on to, in our, you know, contact center, the main line. But even now, we did not put in, you know, as only option.

Like I mentioned earlier, we put as one adoption. So when, as first one, we want to encourage the unit, so we say press one. But some actually, funny story is because we use one as a member service. Member don’t listen the whole thing.

They just press one. So so we say, okay. You got it. Now they tried it.

I said, oh, what is a weak guy? So they actually move out to talking to agent, then he said, who’s a weak guy? I just met earlier. So those are funny, you know, interesting stories, since we just launched the twenty four by seven in the last two or three weeks.

And then before launch, any product is good ideas, you know, to have a team members buy in and education for team members as well as membership. Right? So Vic, the name, actually, it’s a voter from our team. Say, pick a name.

So we, have a, you know, our team, vote on the name, and the Vic was being picked in at that time. And, we also had introduced the vehicle to add our team meeting. So we had the the presentation to the our team. So get everybody prepared.

So every person can answer question if somebody ask, what’s the vehicle on the phone or on the contact center on the website. Right? So not just the only job, it’s, it’s a contact center where you services. It’s everybody’s job to understand the new service that we provide.

And, and then, again, for member, it’s important to have enough education. No matter how much communication, education you have, it’s never enough. And they are still don’t read their emails. They say, what’s this?

So we make sure not only email to them, we, create a landing page, with every single possible questions they may ask. We create the landing page of what to answer those questions. And we call Vic as one of our team members. So make it, like, personalized.

And so I think that is also helpful for member to give a try and if they have any hesitation.

Yeah. Yeah.

You’re you’re you’re Just wanna add something, John, if you go back real quick on this one.

Absolutely. It’s extremely thoughtful approach, like, you know, the phasing and the whole project management team.

You know, the project management team, as Shrey said, is so crucial.

When we look at the, our client implementation gone well and not gone well, the the single most, important in the top three reason would be having a project manager on the client side. Right? Like, that’s really makes a huge difference. Just wanna emphasize that.

And, also, one of the things that, Sherry, you had a luxury of planning these kind of phased approaches because you’re visionary and started way before the, you were ahead of the problem. Right? Always. Like, for example, some of the credit unions on the call may already have a high call wait time, high abandonment rate.

They won’t have that luxury to, like, you know, do such a phased approach. They may wanna jump on to the problem solved right away. But but you’re ahead of the problem. Many times, you’re kind of able to kind of take a phased approach.

And it’s also about the kind of members we serve. Right, so, you know, some met credit membership are more tech savvy, others are not. Then, you know, turning it on all the way twenty four seven is a good idea, which a lot of our credit unions do. But when when they are not tech savvy, it’s kind of taking a slow a slow crawl, walk, and run approach as you rightly done, is makes a lot of sense.

Extremely thoughtful approach and, huge kudos for that. So thank you for sharing that journey as well.

Yeah. And as Sri mentioned, you know, that that thoughtful approach lead to some really great results. So, Sherry, can you tell us a little bit about what impact, you’ve seen? Like, what is the adoption? How is it working?

Yo. So for the phone aside, the the voice AI, we so for twenty four seven, we like I mentioned, it’s only we launched about two or three weeks ago. So this little bit too early to measure the force, full impact on the wait time or call a bending rate at this moment.

But definitely, we see that it’s helping us, already. But we we will have more data to show us later guiding us to do what we can do to improve that. And, so have a a positive indicators at this point, but I would just, like I mentioned, a little bit too early for twenty four by seven, ideas. But, one example of it’s a security fraud prevention.

We, enable the OTP for all type of questions on the phone. Not only initially, it was like, oh, if you’re just asking for, hours, you know, there’s no, OTP, but you ask a balance. It’s okay. But we enable the OTP for all transactions.

And as long as you have, you know, money related question, so that, helping us, to have a security, and, we try to identify the, three potential fraud attempt, in the first two weeks we launched the application. So that’s also another great, approach for our team members really watching those. Right? Say, you know, and then working with you guys for tools to helping us to monitor those situation.

And then it’s not a channel to try to make sure, bad actors not to utilize the service, you know, try to gain some information here.

Yeah. There’s always that balance as you said when you’re talking about OTP between getting the adoption, but making sure that you have all the right checks in place. So it’s something that every c will approach a little differently, but it’s a really important part of the conversation.

Cool. So switching gears to the chat side, you know, we heard what you were thinking on the AI side for voice. But specifically on the chat side, what were the problems that you were trying trying to solve in that case?

Yo. So for us, right, for all the ones. Right? Website is really a critical tool, for prospective membership or members. Right? Searching for information.

And, by having the, chatbot, on the website, Vic will helping the members to find a quest what they’re looking for quicker. Right? And, basic inquiries easier and the twenty four by seven. Right? They get all those frequent ask questions.

And, earlier, we talked about given interface AI, the the, the journey we have gone through. Mhmm. Initially initially, it was like, oh, we need to fill out the template that this is a those type of question can be answered. And the reason the release of the version five, it’s you’ll go ahead and search everything on the website.

And instead that we tell you, oh, we launched a new product. We need to make sure the chatbot you’re able to answer that quest the product. Right? And, we get example a way I can give you is we launch internal launch, appointment scheduler.

And, unfortunately, where the appointment scheduler is hidden in the website, and I had trouble to find where’s appointment scheduler is. And then when I finally said, okay. Let’s just go to leak, ask, then pop up. Oh, it’s here.

This is where it and then then we direct our website to teams. Oh, we need to move this to the front because it’s a hidden. Nobody can find it. Right?

Now they’re easily to find. But chat bot was able to find a way quicker than I was looking for, initially. So that’s just example of, you know, chatbot can help him with the the members or prospective members, to learn more about your credit union. And know what you can do all the services or product that you offer and, then trigger them maybe a call or go to visit the branch.

Yep. Absolutely. And as you’ve said, that progression and that new technology is just making that that easier. And as those new, you know, releases have been layered on and you’ve seen more of those new features, you know, what kind of impact have you seen? This is a product you’ve had for a while.

Yeah. So so seamless member support. Right? The first one talk about, we have the, interface AI, the, AI chatbot, and, we integrate with the live chat and live texting.

So give the members flexibility. Right? They can stay on the chat where they wanna switch to another channel. And for team members, it stay on the same box where they can serve, you know, all the questions being asked earlier to the chat is transferred.

So they know what question was asking, was not able to resolve, and then now what question is being asked. And, so for the members twenty four by seven, right, to, get answer those questions, frequent asking questions.

So yeah. So hopefully, we’ll prospective members or current membership, easily to get to the information they’re looking for much quicker. And, it’s, again, it’s a serving where member want to be served. Right?

The channel they want to pick. And, so so yeah. But if they like to stay on the website because lots of people, right, they’re busy. Right?

They’re at work. They cannot call you. They cannot go visit you. But, oh, they can do a quick chat on the on the website to gather that something they are looking for.

And or if AI able to answer, they can, talk to a live full chat to person. It’s the same person we actually on the phone side. So really knowledge the agent, member service and agent to to answer those questions. So they don’t have to talk to anybody, and then their question is answered.

Mhmm. Absolutely. And as you’ve been nudging, you know, more people towards using chat, I think one of the really exciting things that you’ve seen is this very high automation rate. You know, our best credit unions are performing, you know, at sixty percent, and you’re performing at sixty seven percent.

I think as we were talking about those ways in which you’ve been really artful with, you know, embedding this in different flows, allowing people to choose to engage, has really given, you know, a a pretty impressive automation rate. And for, you know, for any of those listening who aren’t who aren’t clear on this, you know, automation rate is the percentage of people who, engage with the AI who have their questions answered. So, you know, two and three people who say, hey, AI. I’d like an answer.

They’re getting it, today, and that question doesn’t have to go, to the contact center.

So finally, Sherry, the last question is we’ve gone through your journey with both the voice and the chat side. Could you just share your final best practices for other FIs looking to adopt solutions like these?

Yeah. Sure.

So first of all, we, you know, want to make sure you get, the leadership by you. Right? The senior leadership by you. Knowing AI, it’s, not scary.

It’s something enable us, right, to serve better, to for for us credit union to compete with the bigger banks or other fintechs. Right? So knowing we are not behind on technology, using technology as a enabler, right, and helping us. And, so then form a project community team, right, with a different, team members from all different departments of how critical that is.

Right? And, you mentioned earlier some of the good ideas we have. Those are from team members who on on the phone or, you know, on the reply emails, eservices.

They will have a way better good ideas than others. Right? Their first line, they know what the member like to do or type of questions. So we have those knowledge, cross function department, SMMEs on the project. And, dedicated the project manager, we talk about how important that is, from us and as well as the vendor side.

Clear rider collaboration. Right? Set it cleared, expectations for what you’re looking for as a vendor, when you work on the project and knowing what is the negotiable, what is non negotiable, and then talk about just example I gave earlier. Right?

The voice. If Vampers think it’s fine, but we think it’s not, then, okay, that’s nonnegotiable. Let’s just work on that, get it better. And, maintain communication transparency.

Right? What is not working, working, then re reset, problems. Give them the, feedback, right, to the vendor so we improve, the relationship, right, to make a better Resource flexibility. What we did, we did not say, oh, we hire a group of people.

Just work on this project. Instead, we said, okay. Each department, send us some people, to work on this project. That’s a former project team.

So they still have their daytime job, but we know we need a little bit flexibility.

Meaning, certain weeks, we need more hours like a fully testing week. We need them more. So we got the buy in from the leaders from those department.

You know, loan don’t ask for more hours for those weeks or even loan additional team members. Like, oh, we need more people to test him. And then we have that flexibility.

And so, so get the buy in from the management, as well. So we talked about earlier. It is coded by me and, COO. So we we are have a a greatest, intention, want to have this project be successful together.

And the rigorous testing. So vendor can do their testing. They think it’s good enough. We may not.

So it’s okay. I said, no. We’re not ready. And so it’s us at the credit union to make a decision.

I’ll say this is ready or not. Right? So, so that again, we do really, really good away testing. We you should be I know not with you guys.

We got feedback from other vendors that, you know, why you guys find the worst problems so that we actually do is, you know, I said we do. We want to really test all possibilities. Right? So remember, we’ll be now the the first one finding those and they tell us.

We want to be finding and solving those before we launch. And then communication, education, we’d mentioned a little bit earlier. Right? Have a team all pumped up, say, team member excited about work on this project and, knowing this is something going to serve our member.

And, so they got excited. And then for communication to the membership is have those, you know, a dedicated landing page or whatever that may be to helping your membership to be educated. You know your member, so whether which channel. Right?

Whether it’s, so emails or social, you know, you know your member best, what channel is the best for them to learn the knowledge and get do that. And then gather feedback afterwards. Right? We are still early on for the twenty four by seven.

But we when whenever the, they transfer to agent and then agent, they some, you know, they share or our we will be asking those question. How you experience the VIC? Right? Tell me us what we can do to improve.

And, so we we we gather those information as well, and then we can then pass on to interface AI’s team to see what we can improve on. Anything we can do. So those are the best practice I have.

Awesome. Sherry, thank you so much for sharing. As Shree said, you and your team have been incredible partners through this journey, and it’s been just a real honor to to work with with you and your team. So thank you for being here today. Shree, Jack, pass it back over to you for the last few minutes.

I just I just wanna say, if we can make Sherry happy, we can make anyone happy.

Hold us to a high bar.

Yeah. To a high bar.

Okay. Thank you, Sherry. Thank you, John. Okay. So let me actually I have few slides, but I’m gonna skip them. Turn it over to Sri to talk about the evolution of AI.

Absolutely. One second here.

So, I think, Agentic AI is really, gonna fundamentally change how we all go into bank.

Right? So, it’s extremely fascinating time for us to be at this moment, you know, taking advantage of technology to fundamentally change, you know, the the the member experience and employee experience.

And there’s gonna be a really huge positive impact. You know, AI be fully becomes autonomous just like how, cars are driving from, you know, point a to point b autonomously.

We’ve entered the era where AI will drive member experience and employee experience autonomously.

Right? So, it’s it’s a it’s a pretty massive, benefits from that. Members being able to achieve financial well-being if that is their goal and AI talents regarding them through that. Or employee, you know, achieving a complicated task if that is a goal and the AI panels are getting them there.

That’s what agent AI is all about. It’s pretty powerful. And kind of Jack showed real quick on our vision, like, how we are accomplishing that because, ultimately, you wanna have a one AI brain for your members and employees that connects with all of this, you know, technology investments you already made, like LOS, core banking system, RPA system, knowledge management system, CRM. So you let this one AI brain connect and and to these systems.

And now this AI is smart enough to use those integrations to serve your members as well as serve your employees. Right? So you you you set up once, both of them can use. For example, your CFO could be asking, hey.

I need to sell thousand more loans, auto loans in the next one quarter. How do I achieve that? The a imagine AI with the employee AI helping CFO immediately being able to go to the core, understand what is the white space you have with your member base today, what is the propensity of some, you know, certain members looking for a auto loan in the next one quarter based on the transaction history and and who is already actively considering and all of that? And being able to give a plan to your CFO saying, hey.

You know what?

You know, there is about two thousand members potentially can buy or or get a auto loan in the one next one quarter. But by the way, if you can, you know, reduce your, you know, interest rate or give some sort of a discount for members who have two, three products by, let’s say, you know, quarter quarter basis points, your auto loan will be more, competitive compared to your competition nearby.

So do you wanna consider doing that?

Right? With with a click of a button, a c a c f o can take their action and it creates a new auto loan product. Under the member facing AI, you uses that and starts promoting in every interaction. Your members are having with AI.

Wouldn’t that be powerful? You know, you’re putting the entire retail banking on autopilot. Right? So that’s where AI is gonna be going very soon here.

You know, with with that, like, you know, there is just tremendous amount of innovation we are doing. We’re creating a a brand new suite of products that fundamentally changes banking, Sphere and Nexus.

You know, we do have a short period of time. We’ll probably open up for q and a, and we’ll send you a recording of this video of this demo. And I want you to strongly take a look at this. Videos in the Sphere especially is a complete online mobile banking replacement.

It’s a it’s a bank GPT, like a, like a chat GPT. And Nexus is an entire contact center replacement, which is fully AI first. So we’ll send that videos out to you right after the webinar. We got about four minutes,

and there are a couple of questions.

We should, probably address them right away.

I see, Jack, do you, there’s a couple of questions here. One was Yeah.

So I’ve been yeah.

Tried to answer some also regarding us. Just looking at those right now.

Yeah. So, before we ask to answer the questions, just wanna make sure, everybody can kind of, aware of this thirty ninety day free trial. So what she showed you on the UNC site, we can enable the same experience for you on your website, live either on your live website or on your, kind of the testing website. And, you can scan this QR code and sign up for this ninety day free trial.

With that, let’s turn over to questions and answers. So the first question, is for Sherry.

Looks like you answered them already on the on the but let me go ahead and ask you, Sherry. So what are the what has been the impact on your basically, the workforce before and after AI? Were you able to kinda reduce the number of people answering the calls?

No. We did not. We have a thirty or thirty five agent, working, at the contact center.

We did not, reduce anybody. That was the now the intent of we, you know, implement this to replace anybody and, talk about earlier. Right? You want to boost the morale to say we want to have AI.

It’s, we have long call queue sometime and then long waiting time of a higher bending rate. So by having AI, it’s reducing those and then the team members can serving the member better, serving more complex questions. So with that intent, we never said, okay. We’re going to re replace everybody.

So and also we are already on the twenty four by seven.

We may I just could say I have people help can we help them with here? We may say, okay. Maybe not hire as many in the future if we’re blowing up with the more calls, but that just bigger may. Depending on what the calls we come in and then the membership grows. But, yeah, that was not the intent.

Okay. And I guess, Sherry, it’s too early to have a impact on the call volume as well. Right?

Yeah. The call volume, yes. Not too early, to say the changes.

That is, we this twenty thousand calls per month. It’s totally worth receiving the calls. It’s a we lots of members, right, like to talk to us. So, yeah.

So that is a call. Doesn’t mean it’s not being handled by the AI. Right? Total cost come in.

So gradually, we’ll see more those twenty thousand go to AI versus the agent.

So Just, just to provide a little more context to that, Cheyenne, for thanks for the question.

You know, eighty percent of our customers who are on voice AI, who use voice AI, are above fifty five percent call automation. So just gives you their context. Sherry and team are in early part of the journey on the voice AI.

They’re they’re they’re gonna get there, but, eighty percent of our customers are automating fifty five percent of calls or more, through voice AI. Just wanna give you their context.

Okay. So we have one minute, question time for one question, one last question.

Shree, this question for you. How long does it take to integrate your AI with core and online banking?

Yeah. So I see the core that NAS is talking about is Semitar, Episys.

We are already a VP integration partner. Either you’re on prem or on East Cloud, we have VPN tunnels already established to East Cloud.

You know, the fastest, go live we had with Semitar is Great Lakes, which went live in six weeks.

But on an average, it’s about eight to ten weeks.

Got it. Okay. So ninety minutes, is, done.

Looks like thirty minutes. So so, yeah, we covered a lot. Thank you, Sherry, for your time. Really appreciate your insight.

Thank you, John. Thank you, Sri. Really appreciate your time. And thank you for all for attending.

It is so it was very long for probably ninety minutes from your day. So really appreciate it. Please, sign up for demos or free trial. Happy to engage with, you on your journey to AI. Thank you.