Inside Affidian Credit Union's Multichannel AI Strategy for Members and Frontline Staff

Good afternoon. Thank you everybody for joining on this webinar.

We have a special guest from Afidian, Christine Wiley. She’s the president and CEO of Afidian.

We also have Srinivas joining us. He’s the founder and the CEO of Interface dot ai, and I’m the host of this meeting.

Also, we will have Hari joining us to show us a demo of some of the products.

So the agenda for the webinar, I’ll quickly introduce Interface dot ai.

Then we’ll talk to, Christine about her experience with, our solutions.

And then she will come in and, talk about where the AI is going.

Kind of, ultimately, AI is going toward the agentic AI. And we’ll show you the agentic AI demos as well so you can get a feel for, where the agentic AI is going. And then we’ll open it up for q and a, and, we’ll be more than happy to take your questions.

While the webinar is going on, please feel free to type in your questions. Some of the questions we’ll answer right in the in the q and a section, and some of the questions we’ll reserve and we’ll kind of answer at the end of the webinar.

Okay. So before we get started, let’s get a feel for where everybody is in their AI journey. So, Kate, if you can start the quiz.

I’ll give you a minute or two to fill out the quiz. So we are basically figuring out, where you guys are on, your voice AI journey, chat AI journey, and employee AI, whether you guys are deployed it currently. And if not, what is your plan? Like, is it something you wanna do this year, next year, year after next, or no plans at all?

So there are six questions.

Appreciate you taking the time to let us know where you guys are in your journey.

Okay, Kate. Maybe we can see the results.

Okay. So it looks like, majority ninety four percent of people don’t have voice AI. So, hopefully, what you hear and see today will give you a good feeling of what voice AI can do for you.

And then on the chat, fair amount of people have already deployed chat, so that’s kind of my experience as well. I see a lot of chats on different websites.

Employee AI, let’s see where we are. Yeah. Employee AI. Yeah. This kind of fits with, kind of my gut feel as well. Measure the people are very early in figuring out what to do with the employee AI.

Great. So you’ll learn a lot from this webinar. You’ll see the demos. You’ll understand how Christina has deployed it. So this will be very, useful webinar for you guys.

We I have a leaf blower guy outside, so hopefully you guys can hear me.

Okay. So let’s get started with, about interface dot ai.

We have been in the business for almost ten years, and we are focused only on credit unions and community banks.

We are kind of mission driven. Our founder comes from the community, banking background in India. We call them corporate there. And, he has brought his and his family’s experience to basically build the solution for the credit unions and community banks.

We have about hundred customers using our solutions. You can see some of the logos here.

Pretty much we have customers in every state, and, they have been successfully using our AI voice technology, getting, kind of, automation in the forty percent to seventy percent range, depending upon where they are in their journey.

And then we have customers who are using AI chat and AI voice as well. So as you can, kind of, get from my commentary, we have kind of three major solutions. One is AI voice agent for call center. The main idea there is to automate the calls. There’s more to that, but that’s the main idea. AI Copilot for website and mobile app. So this is where, on your, website and mobile app, you can put, the AI chat agent and it can answer very common questions coming from the customers.

Then similarly, we have a AI Copilot for the frontline employees.

So this is either your branch employees or your contact center employees being able to query, and get answers immediately, to the questions, the customer might ask either on the phone or in the branches.

And then we are also working on the agentic AI solutions which is, Orbit, Spear, and Nexus. That’s the kind of the brand name for our solutions, and we’ll show you the demo of that as well. So we have been kind of solving the, providing the solution to the credit unions for, about ten years. And, we have this kind of data advantage.

Right? So we have we have served sixteen million members. So we have heard, sixteen million members saying the, same thing in very different ways. And we, we have understood that and we have kinda learned from that.

So we are able to kinda handle those kind of transactions much more effectively than most of our competitors. Right? Because we have been doing it for longer. And we are kinda doing one point five million conversation daily, which is again giving us this data advantage to understand the context, of the people, in our industry.

The accent, the the different difference where they say the same thing, and so on and so forth.

So if you’re to kind of look at it visually, we we have built a solution on the agentic banking platform. So this is where all the, learnings have gone over the ten years. So a banking platform provides the services to our solutions. This is where we have this knowledge graph.

We have the, kind of the guardrails to make sure we are answering the questions properly. We have all the, kind of the, the the data we have processed. We have learned from the data. This is where our the, the understanding, of the language kind of sits.

And then we surface it to, all the the solutions we provide.

Underneath the banking platform, we have integrated with, pretty much all the major cores and LOS and so on and so forth. We have forty plus integrations for in our banking platform. So it gives, majority of the customers a very fast start where we already done this kind of integrations.

And unlike other vendors, we actually have built our own fraud, prevention solution as well. So, biometric combined with the call center, sorry, the call, I’m drawing a blank. CallForensics, combines us together to provide you the fraud prevention solutions. And you’ll see that, as we go along. So this this is something which is kinda comes in built in, and you don’t have to look at third party vendors for that. So it makes it very easy for you to, to, make sure that you have a, kind of, one of the critical issue we have in the industry is, people trying to commit fraud with a fraud solution. You’re a very, very solid solution.

Obviously, the AI can cannot answer a hundred percent of the questions. So the questions which should go to the, to the call center, human agents, you already have done the necessary authentication and so on. So it saves the time when the call goes to the human agents. So again, voice AI, employee AI, and chat AI, the solution sitting on top of it. And then we are working towards spear, nexus, and orbit. We will which are agentic solution. I will show you the demo of this in in few minutes.

Okay. So there was kind of a high level overview of, interface. So let me bring in Christine. So, Christine, please introduce yourself and tell us more about Effidian.

Absolutely.

Thank you for having me, and I I appreciate the opportunity to speak today about our experience.

Afidian Credit Union was founded in nineteen thirty eight by, about seven Denver police officers.

And at the time, we only served Denver police. And over the course of of several decades, we expanded and and took in about twenty six additional law enforcement agencies, but still we were very specific to serving law enforcement.

And, growth presented member growth really was presenting a challenge.

As law enforcement and just people change in general, we weren’t seeing the same type of participation in academy classes and the need for banking. People were already established and, didn’t really feel the need to have an affinity with a a credit union that just served law enforcement. So, we expanded our char charter in two thousand and twenty two, and we now serve state and local government. So quite a large expansion, but we still had the name at the time Rocky Mountain Law Enforcement Federal Credit Union because we had changed that several years back. So, again, that name being very prohibitive to growth because if you might be a potential in a school district or somewhere, you know, the the name law enforcement doesn’t think that you’re make you feel like you’re probably welcome. So, we we embarked on a rebranding journey that took about two years, very long process. It’s very painstaking, this being the second name change for our credit union, and, ended with Afidian.

And that, we just launched and introduced to members March thirty first of twenty twenty five. So we have a lot more potential members than than we’ve ever had. And, you know, our journey with AI started several years ago, you know, when you could have a chatbot, but they were not smart. So you you could program it, and it could only do so much. And it did provide somewhat of a solution more in an after hours capacity of the, you know, basic what’s the routing number and, you know, locations, shared branch, those kind of things. But it didn’t go really beyond that. So, we I think we got in on that pretty early.

But as AI started to advance and we see, you know, generative and, you know, smart learning, basically, machine learning, we knew that that was an option and something that we would we want to participate in in as well. So, our credit union service about fifteen thousand members right now. We have three branch locations.

We have about forty two employees, very small in that footprint, and, but we’re just about just over three hundred and forty five million in assets.

So, our our strategy for AI is really to be very open minded and look at what’s available.

There there are so many things that AI can do, whether it’s, you know, specific to loans or loan approval or, you know, fraud detection.

There’s just a lot. So, you know, we want to look at it enterprise wide for our organization and, you know, to see what kind of efficiencies that we can create, consistencies that we can improve, service both to internal stakeholders as well as members. And I think that being the impetus for really looking at a solution, like Interface. And, with fifteen thousand members, our call volume was never a problem. So I know a lot of institutions, and I know when I was looking, you know, for for a solution here, I talked to a lot of references as well.

And, typically Christine, maybe we can hold off on this, let’s just kind of go through the slides.

Sure. I’ll come back to the, the the point you are making.

Okay.

So so, Afidian has deployed our AI voice solution, AI employee solution, and AI chat solution. There are two modules with that, smart discovery and smart transaction.

So let’s dig deeper and, first show you the demo of the, phone, AI voice, and AI chat solution.

And then after that, Christian, I’ll come back to you and we will talk about, what are the problem you’re trying to solve and and so on.

Sounds great.

Okay. So real quick, on the AI voice, if people haven’t experienced AI voice, the way it works is basically your existing number remains the same.

Once you deploy our solution, you’re kind of replacing your traditional legacy IVR. Legacy IVR is you call the number, you press one, or two and so on to go to different department, and you might potentially have a phone banking system where people are able to basically punch numbers and get, some phone transaction some, banking transaction done over the phone. So that’s kind of the current environment. So looks like majority of the people who are attending today actually have that IVR system.

The AI voice system we provide replaces the IVR, and then it becomes an, a system where you’re naturally talking to the person who is calling in. So the person who is calling in, instead of, kind of hearing this robotic voice and saying press one and two, And then all the calls going to the human agents, this AI voice will basically act like a human, but, obviously, you don’t want to pretend as a human. So generally, you’ll give it a name. In the case of a FIDN, they call it Aspen. And when you call Aspen, or call the person with AI voice, then the the the phone will be picked up and will say, hi. I’m Aspen. I’m the voice AI.

I can answer, most of your question. How can I help you? The customer will say, what is my checking account balance in a natural way? Right? We our AI understands that the intent is to get the checking account balance.

We are able to, basically go and make a call to your core, figure out what the, balance is. Obviously, we do it after authentication.

Once we authenticate the user, we figure out the, what the balance is. We give the response back to the customer, and they say, how else can I help you? So you’re continuously convert conversing with the caller in a natural way, and Harry will show you the demo and it’ll become, kind of, more, clear on what the AI voice does. Then the other solution is AI chat, and AI chat is sitting on your digital properties. It can be either a mobile app or your website, and it serves four different functions.

The first function is, smart discovery. So this is basically AI powered search. So anybody coming on your website most of the people don’t want to kind of navigate the website, and find the page and read the page.

Everybody’s getting used to the idea of chat GPT. You go in, type the question, get the answer. So this is exactly what smart discovery does. The person comes in, types a question, gets the answer. And it is all done with the generative AI, and Harry will show you the demo of that. Then the smart transaction is similar to your phone banking.

It is but it’s designed for the digital. Right? So as part of the, kind of the natural interaction, I might come in and say I’m looking for a new loan. And then after that, you might say, okay.

Let me figure out what my balance is. Right? And then you can do that smart transaction. We authenticate the user and do the smart transaction.

Then so if somebody likes the the, the product or service you’re offering, which they discover through smart discovery or through navigation, they might want to fill out a form to get the, product, to acquire the product. Right? The loan or whatever. So they are to fill out a fairly long form, and the smart conversion basically helps fill out the form.

So every field will have a helper saying, hey, this feels this is what you need to do. So think of this as a cobrowsing twenty five seven available on each of your form, and it increases the conversion rate. People who might not understand what is the VIN number, why do I need to provide Social Security number. You’re providing the assurance and the guidance to fill out the form.

And also it helps you to understand where the drop offs are so you can kind of clarify what needs to be filled out. Right? Then, the upsell cross sell, the idea there is, somebody’s kind of asking, what is my account balance? If it is very low, then you can say suggest, do you want to kind of have the the overdraft protection and so on?

So in in the natural flow of the things, you can suggest the right upsell and cross sell to the users. You can do that on the phone. You can do it through the AI chat.

Okay. So with that, let me turn over to Hari to show the demos. So let me stop sharing. And, Hari, if you can share your screen and do quick demos.

Yeah. Absolutely, Jack.

Let me share my screen. So we’re gonna start with the voice AI solution. And, just to highlight a few, key features of our voice AI, Jessica is our demo bot. You know, and our voice AI will pick up hundred percent of your calls on day one, essentially eliminating wait times for your members, and customers and reducing abandonment rates as a result of that.

Very quickly, we’ll automate and resolve our voice. We’ll you know, a significant number of the incoming calls, forty to forty five percent within the first month. But our goal, of course, is to take you all to higher levels of automation, sixty percent and more over time.

And then the voice AI is available twenty four seven, nights and weekends as well. Right? So eliminates the need for expensive after hours services, overflow services, you know, while driving consistent experiences for your members during hours as well as after hours.

So let’s, call Jessica, our voice bot, and see what that interaction looks like.

The digital assistant. Please wait while I read the instructions. This call may be recorded for quality assurance.

For best experience, avoid using speaker mode.

How may I help you today?

Hi, Jessica. I’m driving around, and I need to make a deposit. Are there any branches open?

Yes. We are open. Our main street branch only has drive through capability today.

Do you want me to send you driving directions to the nearest branch?

Yes. That would be great.

Great. Please provide me the ZIP code or city name.

I’m in Raleigh, North Carolina.

Thanks. Looks like there are a few branches that are open in your location.

The branches are open from nine AM to five PM, Monday to Friday, and ten AM to two PM on Saturday.

I just sent you a text with driving directions, and you can always talk to me whenever you need help. I am available twenty four by seven.

Is there anything else I can help you with?

Yeah. I just started a new business, and it’s not doing well. I don’t think I’m gonna make payroll. Can you help me skip a loan payment?

I can surely help you with skipping a payment. I have sent you a text with the link to the application to skip a payment along with the instructions.

Please click on the link and follow the instructions.

Please note that the application should be filled at least five days before the payment due date.

To assist small businesses like yours, we have removed the thirty five dollar fee associated with skipping a payment. Do you wanna know which loans are eligible to skip a payment?

No. But can you tell me the balance in my checking account?

Sure. I will help you with your balance.

I see that you already have registered for device biometric.

Would you like to verify using your device?

Yes. Let’s verify using my device.

Great. I’ve sent you a link to your phone. Click on it so I can verify your biometrics.

Thank you. I verified your device.

The balance in your checking account ending with two two three four is two thousand three hundred and thirty seven dollars and fifty five cents.

Also, I wanna remind you that if you maintain a minimum balance of three thousand dollars, you will be eligible for an additional one percent cashback.

Do you wanna know the balance of your savings account?

No. But, I’m traveling to South Africa next week. Can you put a travel notice on my account I mean, my card?

I can help you set a travel notice. I’ve noted that you’re traveling, and I have successfully set a travel notice on your card. We’ll make sure your transactions run smoothly. By the way, you’re traveling a lot, and I recommend you switch to our new credit card with no foreign transaction fee and many more benefits. Do you wanna know more?

No. But my wife is looking to buy a car. Can you help me with car insurance?

Sure. Happy to do that. Give me a moment. I will transfer the call to our specialist, John.

Awesome.

So we saw a few different capabilities in action there, with our voice bot. The first was informational ability to provide informational, responses, where, you know, if I want a routing number, branch hours, location, the bot is able to provide that information.

But, if I want, you know, a behind the pin answer, to an inquiry, like account balance, the voice here is able to authenticate me as a member through, one of a number of different approaches that we enable, and provide me that information. Right?

And then when I asked to put a travel notice on the account, if you all noticed, I said account and then corrected myself to say, I mean, my card. Right? So the bot is able to understand, those kinds of corrections that your members will naturally make as they speak to a human being. Right? And they can continue speaking like that to the voice bot as well. And the bot, understood that accurately and, you know, put the travel notice.

But, it also noticed that, you know, I travel frequently internationally, and it took that opportunity to cross sell, upsell a zero foreign transaction fee credit card to me. Right?

And then, you know, lastly, when, I asked about the car insurance, it, determined that it needs human intervention. It needs help from an agent. So it did a very warm handoff to an agent so that an agent can assist me, understand my needs for the car insurance, and then guide me down the right path.

Very natural voice as you all heard, human like voice. We have about, you know, ten to fifteen different voices out of the box, similar experience, male and female.

We’re also able to train custom voices. So if you wanna train, the bot with the voice of one of your team members, we are happy to do that and able to do that.

So that was our voicebot in action. Now switching to chat AI, as Jack shared, you know, we have a number of different modules that are available.

And we’ll cover the discovery module that we call smart discovery today. Right? So smart discovery really sits on your website, and digital banking, online mobile banking, helps your members easily find products, easily find information that they’re looking for.

But we also have, AI cobrowsing bot, smart conversion.

The smart transaction capability similar to what we saw on the voice AI is available on the chat AI as well, as well as the cross sell up cell module that, again, Jack talked about.

So, let’s, look at, the smart discovery chatbot.

And I’m on, the website of Groove Financial. Groove Financial is one of our customers.

And as we see here, the chatbot is sitting right here on the website. Right?

And, I can start, you know, asking questions just naturally just like I would, you know, just like I did on the voice AI as well, or at how I would speak to an agent. By the way, the bot is trained on all the content on the website. We can train the bot on not only your website, but then also any other member facing content. This is a fully generative AI voice bot. So let’s ask a few questions and, see how the bot responds.

Let’s see.

I have a low credit score. What cards do you offer?

Now, as you see here, when the bot responds, it’ll tailor the response to exactly how the question has been asked. So as you see, it says, if you have a low credit score, you may consider the Grower Visa preferred secured credit card. Right? So, this tailoring is happening at one time, based on how your members are asking question. This is not a static configuration of a response. Right?

And it’s recommending the secured credit card for somebody who has a low credit score. Now and then, of course, it’s providing a lot of details around why it believes, you know, that’s a good card and, you know, the main features of the card as well. But if I scroll down and click on the view more over here, I’m taken to the website, and the bot automatically scrolls and takes me to the right section of the website as well where the product information is available. And it highlights the relevant product as well as you see here.

Now if you look at, the content here, you know, what it says is if you are looking to build or rebuild credit, this is a great card for you. Nowhere in the description for this product does it say, you know, this is for somebody with a low credit score. Right? So what’s really happened is our chat AI has learned and consumed and understood the semantic meaning behind the content on Grow Financial’s website. It’s not just doing a text based match.

And based on the question that the member is asking, it’s which was about low credit score, it’s able to find the right card even though it says that this is for somebody who’s looking to build credit. Right? So it’s able to do that, complex matching, meaning based matching, semantic matching between the question that the member asked and, you know, the answer and the right product that it needs to point them to.

So let’s let’s ask, you know, another question here.

Recently, I decided to buy a restaurant.

Any loans I can get?

No. Again, you know, on rules, let’s say, there is no content specifically around, you know, products for somebody who’s looking to buy a restaurant. Right? But the bot is able to understand that, you know, buying a restaurant is a business, venture.

Right? Business endeavor. And it’s providing various business related products, commercial real estate loans, SBA loans, line of credit, equipment loans, and so on. And then it’s saying consider discussing your specific situation with a loan officer, you know, if you want to, you know, further deep dive into any of these.

Right? And, again, this is a great example of where you cannot control how your members will ask a question. Right? This individual just asked, you know, what do you offer for somebody who’s buying looking to buy a restaurant?

They didn’t ask for what business products do you offer. Right? And the bot needs to be able to understand the questions the way the member asks them and be able to provide the correct answers.

And that’s really the power of advanced fully generative AI, chatbots.

Let’s ask one more question here, and, this will be around, you know, kids’ products.

I’m gonna help my child get started. What do you offer for kids? And as you see here, by design, by intent, this is a very vague question. Right? I’m not asking for specific financial products, but, notice how it’s tailoring the respond response. You know, to help your child get started with financial literacy, we offer the following options.

And it talks about youth savings accounts, activities for learning, you know, and then provides the links to, various products as well.

Yeah. Thank you, Harry. Anything else you wanna show or should we move on?

No. This is it, Jack. Over to you.

Okay. Great. Okay. So let’s continue, with the next topic here.

So she showed you this very compelling demo of AI voice and AI chat. So this on the AI voice, I just wanted to kinda highlight some of the capabilities here.

The voice, as you might have heard, is very, natural, like a human voice. We also support custom voice. So the use case here is if you have a kind of a brand affiliation with a local celebrity, you can use the voice of the local celebrity and that voice can greet the customer and basically be your personality of your, voice AI.

Or you can use a voice of an employee, which is a very good voice and so on. Right?

Better than human understanding, the ASR, speech recognition, automatic speech recognition.

Again, our data advantage, we understand all the financial context and really, like, when somebody says checking, we know it’s checking irrespective of the accent instead of saying chicken. Right? So some of the systems I have seen, don’t even understand the checking properly. So our ASR is finally tuned for our industry.

The device biometric, again, a very effective way to authenticate the users.

Very simple to use. Everybody’s used to using iPhone and doing face recognition to get in. And this is becoming the industry standard, and it’s kind of built in into our solution.

Caller ID forensic to make sure that people are calling, are not, trying to commit fraud. So this is a great defense for that.

Obviously, we need to transfer to the humans, when the AI cannot answer. So the idea is to do a contextual transfer so that when the human picks up the call, he the human can continue from where the AI left off. So the AI is basically giving the summary, giving the sentiment and so on, and giving the status of the authentication and so on, and then passing it to the human agent. And human agent instead of saying, how can I help you?

Say, I see you did this with AI and then continue from there. We can handle complex interaction like joint and trust accounts, upsell and cross sell we talked about. And the analytics analytics, obviously serves the use of the usage, but also it provides you what are the question AI is not answering right now. Right?

So you can say, say, okay. Here’s a question we are getting about cards, and we haven’t done integration with our card system. Maybe that’s the next integration we should pursue so we can cut off we can kind of automate five percent or ten percent more calls. Right?

So this is where Harry mentioned you start with forty percent and get to sixty percent plus is generally driven by more integrations with your kind of back end systems.

Then real time interruption. So similar to, like, IVR. Right? When I when I’m calling the business, which I’m very familiar with, and I’m familiar with the IVR menu, I’ll call, press two, then I’ll press eight. Right? Because I know this is what I need to do.

Same thing will happen with, your, the voice AI.

A lot of peep people will know that if I ask what is my account balance, the AI will obviously answer the question in a full way. So your account balance is this. How can I help you? But if you’re if you know this is, kind of the general response, you don’t have to wait for the AI to complete. As soon as you get the information you’re looking for, you can ask the next question. Right? So it makes it kinda more natural, in interaction.

Then the initial response. So as soon as the person is talking, we are converting the voice into text and and, then we are kind of figuring out the intent and and answering the question. So it makes it fast to answer the question.

So the kind of the tap tap noise you hear when the AI is thinking, that either goes away or reduces.

Obviously, you don’t want the radio silence. So depending upon how far the response is coming, you you might hear one tap or two tap rather than five taps before the answer comes in, or you might hear zero depending upon the speed of the answer. And then, of course, a lot of questions require a text response. For example, what is my routing number?

You can read the routing number, but it will be very nice to get it in a text. So it is saved there for the future use. What is the location of the ATM? Send it to our text.

Right? So combining text and voice, you can provide more effective answers.

Then on the AI chat, as you saw, you’re basically chat jeopardizing your website. So nobody wants to surf the websites anymore. They want answers.

And this is the great way to get answers, and as Harry mentioned, we are learning from your website. Then phone banking without the phone. So everything you can do with the phone banking, you can do it on in the AI chat as well. Right? So you can make it part of the natural, experience where people sometime you’re discovering, sometime you’re doing transaction, everything is happening in the in the one AI chat window.

Smart conversion we talked about. Right? Device biometric applies at AI chat as well. So we can send the link, and then we can authenticate the user. Upsell and cross sells, we talked about this concept.

Escalate to phone. So we our thesis is AI chat is powerful enough that it will automate eighty percent of your calls. And then for the rest twenty percent or so or ten percent or so, do you want to put live chat, or do you want to basically transfer them to the voice?

And the reason to transfer the voice is, like, those are complex interaction, and they might be better handled with the voice call rather than bringing the live chat agent. Right? So if you kind of agree with the thesis, then you can see the implication of this. Right?

So you don’t need a separate a live chat system. You don’t have to pay separate money and separate resources to handle live chat. You can say, okay. Let’s really go big with AI chat, handle pretty much all the interaction, and the complex one are ultimately getting to the phone anyway so you can bypass the whole idea of going to the live chat.

But, again, it’s optional. We do integrate with Glia and L Tropy from the chat perspective, but you should think through whether you really need live chat or not. That multimodal UX idea is that when you give the answer, you want to not only not give the answer always in a text form. You can bring up the UX, for example, credit card.

You can show the credit card as a carousel and people can select, kind of, browse through it and figure out which credit card they want. Live chat integration, we integrate with all the live chat systems, out there. And then again, similar to voice AI, we provide, very compelling, analytics, as well with AI chat.

Okay. So let me, kind of quickly, point you to if you wanna try the FEDN, SPIN, you can call this number or you can go to SPIN’s, FEDN’s website and you can interact with SPIN, on the website, over there.

So, Christine, let’s continue. Sorry for cutting a shot over there.

So That’s okay.

Yeah. So what are the problem you were trying to solve for your members and the credit union? And you’re kind of answering that. So let’s continue with that.

Absolutely. So, I mentioned call wait times, call abandonment. Those are typically, I think, some of the number one priorities that financial institutions will will try to solve for. And and we didn’t have that problem, with the size of membership that we have.

And we have a, a contact center with four individuals. The call volume split over four individuals, including the ability to handle chat at the time was was not problematic. But we knew with the expanded charter and the potential growth, once we change the name, that may change. And we didn’t want to end up, you know, having an issue with service or abandonment times because we, from the member’s perspective, may look like you changed your name, and now we can’t get through on the phone and, you know, service has gone downhill.

So we wanted to identify a solution that would help us, before we actually had any kind of an issue. So we wanted, something that would be scalable, that didn’t require that, you know it’s hard to onboard new employees quickly and have them know all the things they need to know.

But you can absolutely put an AI solution in and be able to ensure efficiency and consistency.

And, you know, again, I I touched on improved absolute improved service for the members, but for the employees too. Because, you know, if you’re using the the frontline assistant in the branches as we are, we have integrated that solution so that within we’re we’re a Scimitar Credit Union. So within the experience module, they can actually click on the frontline agent right there. They don’t have to try to, you know, go to a new link or or find something. I think the integrations are are really important, you know, to the degree that we have those available. So, you know, we were just trying to to be ahead of any issues and not and and not allow a problem that we have to solve for.

Yeah. Christine, the next question I have for you sorry. You’re saying something? No. Okay. So how did you go about the partner selection process, and where did you look for the partner, and why did you go with us?

Well, oftentimes, you know, we we will utilize the vendor halls at conferences. We you know, there are key key conferences that we attend, and, we really use those as an opportunity to search and, you know, seek providers that that we’re trying to identify. So if if we’re looking for something in this realm, then we will utilize those and and then start kind of our search and the due diligence process, references, and demos, and and more demos, and more conversations and and those type of things.

Okay. And the reason for choosing us and, what did you kind of find that tricky about it?

Well, I think the history, the years of experience, the number of engagements that you have every day with consumers, the foundational out of the box solution based on the business that’s already been built, I think is a leader in the market.

The the other part of that that I I I believe to be true, especially now on this end, Interface, you you you all have, you know, become really good business partners, and I think that’s always true typically on the sales side, but on the support side, it’s more important. And, you know, you can see from, you know, when we implemented, we’re still, you know, still brand new, but we still have, you know, a weekly meeting. And, you know, I feel like you’re you’re right alongside us trying to if we identify something that like our Aspen, we we don’t like the look of Aspen.

We think it would be more friendly for our members if it looked more human and less robotic. And to be able to partner and come along and give us some options and solutions, it it really is an indicator of a true partnership because it’s a win win.

Right. We want to be able to be a really good customer and and a voice for interface because you, you know, are are doing that and and such a good partner for us. So it’s it feels like a good, you know, mutual engagement.

Yeah. Yeah. And Christian and I are actually working on something cool, using the latest, v o three technology to kind of bring, humanize the whole, AI voice. So that will be coming soon on their website.

Okay. So, obviously, implementation.

This is one of the, thing which worries the credit unions the most. Like, how complex is the integration? How do we go about structuring our team? So can you please talk to that, Christine?

Yeah. I think that’s critical point, with any project is to identify who needs to be internally on the project and to make sure externally on the other side that that you have strong players as well.

You know, not not in this example, but we have had projects where we get someone on the the opposite side that’s like a brand new project lead. And, you know, those kind of things aren’t conducive to a good start, and and that’s not been the case at all with Interface. Everybody’s been fully engaged, very knowledgeable. And internally, we have a very strong team, but a very cross departmental departmental team. So very high IT participation.

We have our trainer participating. We have the head of our digital branch participating, and then we have, two other executives participating. And, you know, these are weekly meetings. There are often in between meetings, assignments, and you’re, you know, you’re filling out things that need to be filled out so Interface can take it back and build it. And, you know, so make making sure that, internally, those teams are staying on timelines and, because our implementation was quick. Jack, would you say? Agree?

Yep. Mhmm.

Yeah. Very quick. So I think, you know, you you have to identify how big of a priority this is as an organization and put the the required resources necessary to get there. But but other than that, it’s it’s been a great project.

Great.

Okay. So moving on.

Could you share the best practices for FIs looking to adopt the solution? So you kind of answered that question a little bit, so let’s just go a little deeper.

I I one thing I would say for sure is, well, that that’s a great lead in, is is definitely buy in. Right? Is to, you know, have conversations, you know, before it’s just, you know, the present sitting on someone’s desk. Right? So that that people get an opportunity to participate in decision making, understand the need, and the the app the potential benefit.

And, I did mention making sure you have the right internal team.

Make sure you find the right partner.

Absolutely. Communication is critical. In between our weekly meetings, there are numerous communications that go back and forth. So it’s not, meet on a Tuesday morning and we’ll talk next Tuesday. It’s a continual based on, you know, things that that are going on or or we may be reading a transcript transcript from an engagement and want, you know, to to change something a little bit, and so we can you know, we don’t have to wait for those meetings. So communication is really strong.

And what else?

Yeah.

So Ongoing commitment.

I mean, you know, we’re months and months in, and, you know, the commitment is as strong today on both sides as it was, you know, even at kickoff.

Yeah. Yeah. So this is so critical. Right? It’s like getting the AI to handle the calls and chat is like hiring an employee.

For interface dot ai kind of gives you the employee which is kind of trained on seven seventy, eighty percent of the intents. But obviously, you need to kind of make it work for your brand and so the questions which are relevant to your brand, which are unique to your brand.

So putting that work upfront, we And ongoing.

Because I think you have to you know, once you have everything up and running, you know, like for us, we have goals and I think, you know, we’re we’re sitting up about forty eight percent. I think I looked, you know, earlier this morning, on, you know, the autumn automation adoption. And, you know, we wanna continue to have that number grow. So it means, you know, looking at the transcripts and identifying things that both sides can do. And we’ve done call out campaigns for members, trying to educate members, and and things like that. So I think it has to be, you know, kind of ongoing.

Right.

Yeah.

Great. Okay. So let’s, switch gears and talk about, the employee AI. Employee AI is basically the frontline assistant.

And, let’s kinda go into the demo. So, Hari, I’ll stop sharing and and I’ll show the quick demo, please.

Yep. Absolutely. Let me, share my screen and, walk you through the demo. So as you see here, I’m logged into, employee AI, and, you know, there’s a call that’s waiting here.

Right? So I click on the call. The agent picks up the call. Now there’s a number of different, you know, pieces of information that are presented to me, as a, MSR.

On the far right here, as you see, the full authentication status, including caller ID forensics, you know, that was applied on the call is surfaced to the agent.

That way, they don’t have to reauthenticate the member, spend another sixty to ninety seconds, on authentication. That’s an immediate direct, savings and a positive impact on AHT.

In the middle of the screen here, the full context of the conversation, the full transcript, you know, the conversation between the AI and the member is also conveyed to the agent. That way they know what transpired and what was the reason for the transfer.

Right?

Now, because, you know, it’s a single unified AI brain that goes across, our member facing solutions and employee facing solutions.

The AI takes the reason for the transfer, goes and searches your policy procedure documents, and finds the right resolution and presents it to me.

Now if I click on view document, I can, you know, verify that it is coming from the correct document.

You know, that way I can make sure that the AI is not hallucinating and, you know, it it is providing me the information from the right place.

So this is a knowledge assistant capability, but our AI goes beyond knowledge assistant.

Now let’s say I want to help a member lock a credit card. Right? So in other words, perform transactional activities on behalf of the member.

You know, we integrate with not only your core on the employee AI side, but also your lending system, card processing system, you know, really all the system that your, agents, MSRs need to swivel chair and work through.

And our employee AI presents that single pane of glass, where agents are just, interacting with the employee AI, and employee AI is then integrating and interfacing with those back end systems.

So let’s say a member wants to lock a credit card. You know, the suspect fraud has happened. They no longer want to use that card.

So all, I do as an agent is, tell the AI to log the card ending in, the last four digits for this particular member.

And, the AI takes care of doing the locking. Right? So it’s going and finding the card for me, and, it’s gonna ask, you know, if I really want to log this card and, not use this card anymore.

So as you see here, you know, it formed that it’s a cash back credit card attached to this particular customer. You know, are you sure you wanna log this card? So let’s go ahead and say yes. And it connects to my card processing system and then takes care of locking the card.

Now and then it’s asking me, you know, is there anything else I can help you with? Now this member doesn’t want plan to unlock this card. They just want to apply for a new card.

So let’s say, apply for a credit card.

It’s as simple as that. Right? Your, your agents are conversationally interacting with Frontline AI. And as you see here, the Frontline AI provided, details on, you know, how to apply for the card as well as, you know, a link to the document here.

But then also what we see down here is a view experience. Now what this is is, you know, for this particular, organization, the member facing AI, the self-service AI, has been trained to take a credit card application just through self-service.

And because, you know, our AI engine goes across all of our solutions, The AI engine is able to take that training and make it available on the employee AI, to enable your agents to take a application on behalf of that member.

So let’s click on this view experience, and, that, triggers that, workflow. And as we see here, it’s saying, sure. I can help you take a credit card. Now it’s gonna coach, me to ask, specific questions of the member that’ll help the AI complete that application.

So it’s saying, please ask the customer what type of credit card are you looking for, and I can suggest, better options for you. So let’s say the customer wants to stay with the cashback credit card. The AI provides the details on the cashback credit card in a very easy to consume fashion for the agent, including all the key features, you know, of the card, and it’s coaching the agent to read it back to the member and, you know, get a response. Right? So let’s say the member is satisfied. You know, I go ahead and click on satisfied. Now it’s connecting to my lending system.

And, it’s gonna, you know, take that application, you know, just conversationally here. Right? So it’s asking me to confirm the customer’s first name, customer’s last name. It just needs a few pieces of information here.

Date of birth. Let’s provide the date of birth.

The, annual income, and then, the contact information.

So let’s provide the contact info as well.

And as you see here, sensitive information is masking. So, you know, it’s, summarizing the information. You know, it’s asking if everything looks good to me.

I say, yep. Everything looks good. So it goes ahead and submits the application. Right?

And now it’s taking the opportunity to present a money market account to me because, you know, it knows based on member behavior that folks who, have in the past applied for this credit card also very likely sign up for a money market account.

So, you know, the cross sell, upsell aspect that we saw on the member facing side, is also available where we are coaching your agents to engage in those cross sell upsell conversations.

Now this is so important because as we all know, we hire agents for exceptional service skills, not sales skills. Right? So having a tool like this and AI being able to coach your agents and assist your agents in presenting these upsell, engagements, becomes a very powerful, utility for them.

So, yeah. So so this is the branch experience or rather the contact center experience. Now if I exit out of this and, go to the branch experience, we do have a branch experience as well for your employee AI, where, you know, let’s say a member walks into the branch and, they are interacting with your member facing staff.

AI is available to them as well, to enable them to drive the same consistent high quality experiences for your members. So let’s say a member walks in and wants to know if there is a specific policy procedure that they need to, adhere to in order to withdraw a large sum of cash.

All the agent needs to do there or MSR needs to do the branch employee is just ask the AI.

Just large cash order. I’m not even typing a full sentence as you see here. Right? And, similar to what we saw with the contact center experience, the AI is able to summarize, you know, the the procedure and policy for withdrawing a large amount of cash.

And, similar to what we saw in the contact center view, I’m able to click on the view document. And, the AI not only opens the document, takes me to the right section of the document, and then highlights, you know, where it got that guidance from. Right? So that I can very quickly verify it is coming from the right source.

And then the last piece I wanna show here is, our knowledge repository and, you know, some of the, related features here. Right? So as you see here, we are uploaded, and you’re able to upload, your policy procedure docs within our platform.

This is not a replacement for a full fledged knowledge, management system. So, usually, what our customers will do if they have a SharePoint and or another knowledge repo, we’ll just point to their existing repo. That way, they don’t have to replicate and duplicate their documents.

Thank you, Harry.

Okay. So, to summarize what Harry showed you, AI conversation summary coming in, gen AI answers for your employees. As Christian mentioned, this is helping her employees answer the question faster.

Automated actions, which we, Harry showed you. Authentication and caller ID status, is borrower. So when the call is transferred to the human agent, they can, save a minute or two. Upsell and cross sell, Harry showed you that.

Third party app plug ins. So this is kind of differentiate differentiator for us. Right? So it’s not just we are answering the questions, but we are making it easy to take automated action into your into your back end system.

So a call center agent doesn’t have to go to seven, eight, ten different application to do things. They can simply click buttons and get things done. So it is a huge productivity gainer there. Knowledge management, you can we can point to your knowledge management system or we can kind of act as one.

And then you can have role based access control. So you can say only these documents are only for loan officer. These are for collection officers and so on and so forth.

Okay. So, Christine, if you’re, back, let’s talk about, the employee AI. Like, what are the problem you’re trying to solve and, with the frontline assistant for your credit union?

You bet.

So organizationally, we had Pat well, we still have, but we utilize SharePoint a lot. We have a shared internal drive as well where there’s a lot of documents, and there’s just a lot like like we talked about fragmentation of where to find data and resources or the the most recent form and those type things. So, you know, when people can’t find things and there’s just too much for somebody to have to retain, you know, without having the resource to go to, you you end up having sometimes wrong answers, inconsistent, you know, or inconsistent answers. And so we would see team teams chats a lot.

Internally, we call it phone a friend. And, you know, some of those friends internally are really popular because they get a lot of internal calls. They’ve been here a long time and have a lot of knowledge, but but it’s just not efficient. And so, you know, much like we’re trying to create consistency for members, we wanted to create that kind of a resource for our employees so they then can provide that, you know, consistency and answers and and support and service to the members.

And it takes time to try to go fish around. Right? So it is a lot quicker when you can use the frontline assistant to try to find something. Forms as well.

So, for us, we we have been uploading. We we just went live mid April, and it it was it it wasn’t as quick quickly taken by employees as I would have thought it would be, but it was another link.

So once we were able to integrate it in right within Jack Henry, and it’s so it’s right in the experience module, it’s right there for them to just have up all the time and and quickly be able to typing type something in just like Harry was showing.

Yeah. So it definitely has been a fast start and, logged by our members, by our employees. So Yeah. Very much. About the impact and the adoption. You kind of briefly mentioned it?

Yeah.

Like I said, since being able to integrate it into Jack Henry, it it’s been an immense night and day, hundred and eighty degree change.

Staff loves it. It’s a great resource. I think internally, it feels like it’s a benefit for them. It it eliminates a lot of frustration and feeling like I’m stuck because I’m trying to provide the service that, you know, everybody wants me to provide, but I don’t know where to go. And so this is just such an easy resource. I think it it feels really positive internally.

And, again, they they know where to go find stuff. It’s consistent. It’s it’s a hundred percent answering questions.

We’re we’re loading forms and and putting things out there, making it really easy for everyone. So it’s very quick.

Okay. Thank you, Christine. So now let’s talk about okay, there’s one more question. I think we’ll skip this. I think it’s the same best practices you had for other projects, I believe. Right, Christine?

Yeah. I would absolutely say input and getting feedback from employees too. That that’s really helpful.

Got it. Okay. So let’s talk about where the AI is going. So, Sri, can you come in and kind of educate us on where the AI is going?

Absolutely. And and, Christian, Jack, and Harry appreciate, kind of showing how this project has been, impactful. The AI has been impactful across organization.

You know, but pretty much, wanted to give you kind of a big picture, how this technology is evolving really fast, what we showed you, today, and what is coming in the future, and, you know, how as a company, we’re kind of having a, you know, vision to drive, you know, more value to our customers, now and tomorrow with the constantly evolving AI. Right? So, you know, if you look at, this graph here, you know, pretty much, AI is moving really fast. Right?

Like, you know, couple of years back, we had conversational AI, which is more like NLP, which required quite a bit of, overhead to manage and get value out of it. Right? And today, we are in generative AI, which is kind of makes everything little straightforward. The AI has ability to, you can point the AI to knowledge you’ve trained for couple of decades.

It learn everything from that in a in a matter of few minutes.

You know, basically helping finding information easier and faster.

But where agent AI is going is, quite amazing possibilities. Right? So, you know, in in kind of a general example I often give is, generative AI is, you know, as a technology, let’s say, you wanna find the best recipe online.

Let’s say you’re kind of making a a Thanksgiving dinner, like, you wanna find a best recipe for us, for this particular dish, you know, you know, generative AI is the one that searches the Internet, finds you the best recipe.

But agentic AI is the one that now goes ahead, finds the, you know, groceries required for you cooking that recipe, and automatically goes to amazon dot com, orders that for you, and make sure the groceries come to your doorstep. That’s agentic AI. That’s kind of difference of generative and agentic AI. In a banking world, and a generative AI, you know, could help, you know, help you find a, product or service, but agent AI could apply on behalf of the user for you. Right? Like, that’s kind of really powerful technology is in the near future.

Our vision as a company is also evolving with this technology. So, if you go to next slide, Jack, real quick. So, you know, in the past, as part of conversational AI, we had our AI embed in each of your member communication channels or customer communication channels. You had online mobile banking, website, text, contacts, and IVR.

We had AI plugged into each one of those places. AI achieved, somewhere between, you know, thirty, forty percent peak, fifty percent kind of automation, in this era. Right? It required a lot of manual effort to train this and set it up and things like that.

You know, today, in the generative AI era we are in, the AI is kind of leapfrogging beyond what it was able to accomplish and going to sixty, seventy, in some cases, even eighty percent call automation or chat automation.

And this AI has now become like, this one AI brain, you’re able to train once and and deploy across the organization or channels. Right? And eventually, the the role all of the, channels would play would diminish, you know, pretty much. But the future, is what we call autonomous AI, which is where kind of bank GPT.

You know, your members are able to add multiple financial institution across, you know, let’s say, you know, today, average number of accounts, someone leaving the US, kind of bank is about eight to twelve different financial institution relationship.

So they could now, eventually, with Agentic AI, they could have a bank GPD from you, you know, as your members and customers, and they can add all other financial institutions and bank across all of them at one place. That’s where we are heading with, Agentic AI. As the technology evolves, you know, we’re gonna create, like, a bank GPT, which pretty much replaces the need of online mobile banking the way it exists today.

And and you you’ll be able to put, like, an AI financial companion in your member’s pocket that can proactively drive financial well-being and things like that. So it’s a very exciting future coming ahead with agent AI.

What we have done at Interface AI is, we have gone ahead, already started designing, building products around agent AI, capabilities.

Right? So if you if you if you look at our current products at the bottom row, which is voice AI, employee AI, chat AI, you know, all of them are transitioning to become agent AI. But also, agent AI is so big that we have introduced brand new products, that kind of, you know, meets the expectation of what technology could do and meets the, kind of the value that our customers are expecting. Like Sphere, Nexus, Orbit, at the top row are all brand new products, you know, leveraging fully agentic capabilities.

Right? So, you know, what I I plan to do for the rest of the meeting is I wanna show you a demo of this new agentic first brand new products, at Interface AI. It’s fundamentally is gonna change how banking is done. I’m gonna, invite my colleague to help me with the demo.

So we’ll start with Orbit, and and we go from left to right. What Orbit does, in in is fully agent so if if you this is a hypothetical bank we created. Don’t worry. We’re not in banking business.

So, you know, if you go to the right bottom corner, you see a chatbot there. You click and open it. So it pops up like a regular chatbot. Right?

You could ask questions, you know, what is my routing number? For example, as how you are showing. It just answers those, questions. Generative AI, nothing agentic here.

But let’s say now, you’re gonna ask a question.

You know, I’m a student looking for a credit card. It follows up and asks you, you know, what any features you’re looking at. So I’m gonna say the feature, and I’m just gonna show you agentic AI in action shortly. So when we type it up, it comes back and says, hey, do you want me do you have any other preference or should I should I go ahead, proceed? So say no no other preference right now. Go ahead.

It takes over the entire website and automatically navigates on behalf of the member and the customer and searches the website to find the right product, that matches their expectation.

Comes back, shows you on the right side, here’s a product. On the left side, it highlights, a a a particular product, that that is best suited for you and and auto scrolls your website there. So as you can see, AI has kind of become a companion to your website. So typically at this stage, if someone is interested in such a product, applying for such a product is a conversion process. Right? There is this multiple steps of forms for a especially a new member.

It it takes a lot of time. But let me show you how with agent AI that is going to change. So here it says, do you wanna apply? I’m just gonna say, yes. Let’s apply.

And, you know, immediately, it asked me, hey. You know what? Upload these documents. I’ll fill up the application on behalf of you. So we are gonna upload those documents.

It is gonna process those documents and come back and let me know if anything else is needed to fill up the application form. It says, hey. You know what? I have all the information to fill up the application form. Should I proceed? I’m just gonna say yes.

And, again, take over the entire screen from the member, Automatically navigating the application form, filling up the details on behalf of the member and the customer.

Right? And wrapping up the entire application, on behalf of them. Right? And gives you the confirmation number.

Imagine the old days where how much effort your members have to put or customers has to put to go through filling up this application form now. They just have to upload the documents. They’ll take care of it, on behalf of them. Right?

So that’s the power of agent AI is gonna where we’re heading next. Right? So let’s say now they’re in online mobile banking.

Right? So let’s quickly log in, and, you’ll also have an AI as Harish showed you, you know, this exists and we, we can try, let’s say, fund transfer or something we, which, Harry also showed you. This is kind of a basic chatbot which is already available, which is all generative capability.

Right?

So you click yes and it takes care of the transfer. Nothing agentic here. Right? But let me show you the agentic part of it.

We are gonna ask something, you cannot do it on online mobile banking today. Right? So we say, what are the streaming sir, service sub subscriptions do do do I have? Show me all the expenses and streaming service in last one year.

So AI is now gonna become agentic and take take over the entire online, banking screen. So let’s click yes.

So it takes over the entire online mobile banking system and automatically navigates, goes through every account, looks through every transaction every, you know, statement, and every transaction within the statement automatically classifies all the expenses into a a really interactive graph to tell us, like, you know, how much the expenses are in streaming service across the year and different streaming services and things like that. Right? So, you know, that’s kind of agentic capability. You know, Orbit, which is a product we just saw, which can help you on the website or online application or on on your online mobile banking, is pretty much is an AI that is trying to snap onto your existing, you know, online investments your existing in investments in technology and make sure they’re enhanced through AI. Right? But where we are going next is, you know, hey. How would a new online mobile banking may look if it was fully AI first?

Right? So that’s what we call sphere. So what do you see on my screen is a brand new online mobile banking system, which is fully AI first.

Right? You can you can see account aggregation. You can see multiple accounts that you already have. On the right side, you have it, in a AI chat that you could, if as soon as you start interacting, it takes over the full screen and kind of becomes bank GPT.

Right? How can I help you? So you can ask extremely complicated questions that you could never imagine online world banking would ever do. For example, I’m I’m gonna ask, I have a ten year old kid. I’m looking to, plan for an investment for their college education.

Right? And AI quickly comes back with a few follow-up questions. Tell me the kind of college, tell me the university, details, what all, you know, accounts you wanna be considered for the planning. Right? And what is the monthly contribution comfortable with? And then it asks, do you wanna consider five twenty nine plan as a strategy? So you can make all those preferences.

You can say, I wanna begin my research.

Right? And it has come it has come back and ask you, hey, you know, is there any other accounts of your family members you wanna add? Like, for example, you may wanna add your spouse account. This is a big decision, big savings.

Right? So you wanna plan together. So you can literally add that right here. So you could say connect connect connect your account.

It’ll ask you which credit union they’re banking with or, you know, financial institution. So it’ll immediately open up.

Here it’s a safe credit union. Immediately automatically load safe credit union.

You know, they can log in right here and, you know, pretty much it it connects to the account and pulls all the, different accounts available and comes back and tells, hey. I got, you know, two checking accounts, a savings account. Should I consider this for planning? You could say yes.

And, you know, you can you can see now as you scroll down, those accounts are added. So you can check them, for the planning and say now looks good proceed for research.

Now the AI is actually doing research how college tuition’s have, fees have changed the last couple of decades and how it is the college tuition fees are changing for the next decade, step by step research, which should have taken you many weeks to do that research. It is doing all those research coming up coming back with a a plan for you, on what is really required for you to open up a, five twenty nine investment strategy. So it says this is the amount you should be planning for, for so and so college. It is the savings you should probably plan for, monthly amount contribution.

And at the bottom, you know, it shows investment strategies, education savings account. Now with the click of a button, which says execute this plan, it has the ability to open all those accounts, set up auto transfers, and set up the entire five twenty nine investment plan for you without you having to go do that from multiple places.

Right? So that’s how the future of online world banking would look like and really excited to, pave the pave pave the path here.

And finally, I I wanna show you one last, new product that’s coming up, NexSys.

NexSys is a entire contact center a contact center replacement. The contact center as you know today would no longer exist. Right? So it’s not going to, you know, the the contact center we know today is gonna pick up the calls or chats.

It has a routing engine. It has a workforce management engine. And then, you know, and and and and then a a agent portal for picking up the call, handling, and things like that. That’s not how contact center would exist in a agentic world in a just a couple of years from now.

This is the screen your agent will be presented.

So as soon as your agent says I’m available, immediately, AI ascends them a task.

Your member is talking to AI. Your agent is also talking to AI. Right? So the the humans are no longer talking to each other.

The AI they have they have a AI in the loop all the time. Right? So here, a task is assigned to your agent. With a real time call or a chat happening, the AI bought in your agent in in the loop asking, hey, you know what?

I got stuck here. Can you help me? I already executed nine out of the ten steps for this wire transfer, but I got stuck here because the session got expired. So all your agent has to do is click here, put the passcode, re enable the session, and it it shows you back a confirmation step agent needs to approve, for compliance reason where the agent could say, yes, authorize.

That’s it. Imagine the wire transfer if your agent has to do all all of it themselves. Instead now, AI took care of this. And nine out of the ten steps, agent has to come for just one step.

Now they are done with it. AI has already assigned them next task. Right? So it seems like the next task for your agent is a callback request.

It seems like the callback request was regarding loan application status. You can click a button to initiate the callback. And if you see on the left side of the screen, the AI already has bought in the right tool and the right screen for your agent to take actions.

So as you could see on the right bottom corner, you can see the call, call happening real time. Right? So, it’s real time transcribing.

Now the agent is done. They’re assigned with a new task.

This time, it is about increasing credit limit.

Right? So here, again, AI has done nine out of the ten, steps, but it got stuck with the last one where it requires any, income statement is too old. So, you know, the agent gets involved, says request the latest income. It seems like member has already sent them to their, support email.

All agent has to do is click a button. AI will automatically read the emails and find out that latest, statement they have sent. Now it now it has approved them. It asked agent to also approve.

Right? So the moment they do, the task is done. Again, as you noticed, the agent is going from one task to another. It it’s different calls.

What you saw is three different calls, and they’re parallelly managing that because AI brings them in at the right time just to unblock where AI got stuck. And it takes over again from them. So, you know, twenty years back when contact center came, you know, you were handling one call at a time. We even still in contact center handle one call at a time.

But then live chat came. Our agents were able to do four to five chats at a time because they’re able to multitask.

But with AI, bringing them only during a a one or two steps in a ten step process, for a call or a chat, your agents, we anticipate will be able to do thirty calls or chats at a time. Right? So, that’s what agent AI is going towards. We’re really excited the kind of productivity and and the improvement in self-service.

You know, if you look at the member who is either talking on a phone or on a digital, system, that is powered by AI, they don’t even know there was an agent got involved to help resolve their issue. Right? So, that’s how kind of, AI brings a human in the loop to complete them. So that’s how the future we’re we’re excited about.

You know, I I know, we’re on to the q and a section now.

Jack, back to you. We’re happy to answer some of the questions you’ve been waiting for.

Yeah. Sure. Thank you, Sri. So I’ve been in the contact center industry for a while, and this is, this is basically the next generation of, customer service. So we’re very, very powerful and amazing, when it comes to the flirtation.

Okay. So we are ready for q and a.

So, obviously, we would love for you guys to see our product. We can do one on one demos.

So if you are interested, you can kind of scan this QR code and book a demo right now. And the first five registrants will actually get, the device biometric without any licensing fees. So the usage fee will apply, but there’s no license fee for the first five who register.

So let’s go to the q and a. The QR code is still here, so feel free to use the QR code to book the demo. So there are quite a few questions we already answered in the in the in the in the chat. And then there are two questions, which I see open, which require answer, from Christine. So, Christine, if you’re around, the first question for you is how many agents do you have, and what are the call volume?

We have four agents.

One of those agents is doing a lot of review of all the the analytics and the transcripts, so he he’s not really full time really interacting.

And, generally, we have about two hundred calls come in a day. And I think I mentioned we’re at about forty eight percent when I looked this morning, adoption.

Got it. And, Christina, the next question is for you as well. Who admin the product now? I think they’re talking about the frontline and that you’re live loading the forms and information for the frontline. Does that sit in IT, training, ops, retail?

The project team is handling that. So it’s the group of, a combination of we’ve got two from IT. We’ve got our trainer, and then primarily, those are the three. And and one of the one that leads digital.

Got it.

Better that’s uploading the forms and those kind of things.

Okay.

Shree, I’ll give this question to you. So the question is, do you sell voice AI chat and employee chat separately or all this is all packaged together?

If you all buy them together, you set up once, deploy, all of them. But if you but you do have an option of a la carte. Right? So you can take just one of them. They’re not fully packaged. But if you do take them, it it they work together, better, but you have an option to take a la carte as well.

Got it.

Okay. So I’m gonna combine two questions, and I’ll ask it to Shree.

What is the estimated asset member size of your smallest client? And there was a question about, what is the largest client we are working as well, which which I answered online. So maybe you can kind of give us as a sense of where our customer base is.

Yeah. Yeah. So I think, our sweet spot has been about three hundred million dollars in asset at at at the least and to all the way, with twenty billion dollars in asset, our customers today. And with with some of our customers’ recent acquisition, now our largest customer is about forty five billion dollars. Right? So, but, otherwise, the kind of sweet spot has always been kind of about three hundred million to about, twenty billion dollars in asset.

Okay.

Okay. This is a really good question. I think Christine and Shree both can answer. Maybe, Christine, you can answer it first. Yeah. So what is the average time it takes for member adoption? Did you find any members extremely upset that they could not get to the real person immediately or not liking the new AI, or was it pretty seamless and taken too well?

That is a really good question. And, again, we we haven’t been live that long. So to have, you know, come on board live with members or Aspen, as we call her, just about mid February, I I think, you know, I’m not disappointed at the forty eight percent because in a small membership where we’re not having issues and members are used to be able to get get a person and not have to hold for ten, fifteen, twenty minutes or, you know, a lot of call abandonment, I think that’s a great number for us, although that’s not our goal. We we didn’t have members upset, and we we took a really member friendly approach when we launched, and that is to still make an agent available as an option.

A lot of I I believe a lot of times, customers will, like, make that option hidden or or not even available and kind of force a member to have to start with with the agent. And and we didn’t you know, it it it to me, it felt like in our membership, at least in our group, that that that wasn’t a service focus that so we so we didn’t do that. So, but we didn’t have anybody upset.

Okay. Sri, maybe you can, answer this cross question broadly across our from customers.

Yeah. I I think, across our customers, as Christine rightly mentioned, these are about you know, our platform is highly configurable because our customers have different preference. We have about two thousand configurations, honestly. Right? So, you know, pretty much you can tune the bot where where you want to push harder for, you know, self-service or you don’t wanna do it. Like, as Christian said, they take on a member friendly approach, which is, the you know, the moment someone is asking for an agent, it is reacting and quickly passing to an agent.

But in some cases, our, customers have taken a little bit more self-service approach where they’re creating a little bit of friction to drive adoption. If someone says agent immediately, it is it is kind of pushing them. Hey. Why don’t you tell me at least a question you’re calling about? So there is different levels of optimizations depending on your, depend depending on your preferences, you can kind of adjust accordingly.

Okay.

Okay. The another really good question.

Maybe, Sri, you can answer this. If an FI were to consider crawl, walk, and then run approach Yeah. What approach would you advise?

Yeah. No. All of our FIs take crawl, walk, and run approach, mostly. Right? So I think, it depends on why you’re thinking crawl, walk, and run.

Sometimes you arrive at the conclusion because your board may not understand technology. It’s just hard for them to sell the technology and things like that. Or sometimes you the the other reason you may think about crawl, walk, and run is you’re worried about the drastic change in membership. You wanna really start small.

It depends on why you think about crawl, walk, and run. You know, in case you’re thinking about, you know, is is a board or member adoption risk, a member experience risk, you can always start from, you know, adding AI for, employee first. Or you can start with the, chat AI, which is on the website. It’s not gonna disrupt, your your, member experience in a big way.

But let’s say you have a big pinpoint in contact center on voice calls and you really want to badly do something there, you you also have options to do crawl, walk, and run even there because you could just roll it out for after hours only. You could just roll out the voice bot only after hours and then roll it out twenty four seven later on. So there’s a lot of options. Either you start an employee AI or chat AI, or if you really need, need a voice AI, you can start after hours.

Is those are kind of the starting point of crawl.

Got it.

Great. So we basically have one minute. And, maybe the last question I can give to Christine, I’ll combine the question.

The question is, where did the question go? I I guess it got answered.

The what is your online banking system? Bano or something else? And what was your, kind of the, automation rate before you implemented interface?

We had ninety ninety eight percent plus probably calls answered because unless a member had to hang up or something couldn’t go through, you know, taking member calls weren’t what wasn’t an issue. Again, for us, this was preparation for an expanded field of membership and, you know, trying to imagine if if you doubled your calls in, you know, six months, you wouldn’t be able to have that same level of service.

So ours again, not answering that problem, but being more proactive. And we are not on Bano. We’re on Alchemy, Alchemy’s platform for digital and online banking.

Perfect.

Okay. Christine, really appreciate you taking the time.

Shree, thanks for the demo.

And, the attendees, if you guys, if you want to set up a demo, please, go to interface dot a I and book a demo there.

Really looking forward to engaging with, everybody here. Thank you so much.