speaker-0: thank you, CJ, ⁓ for taking time of your day to hop on what I think is going to be a informative and fun ⁓ webinar. ⁓ And thank you all ⁓ who are attending webinar. ⁓ As we're going through this, we will leave ample time for any Q &A at the end. So. There is the ability to leave questions in the chat. We have members of our team that are ⁓ reviewing that and surfacing that and CJ and I will be happy to answer that. ⁓ More so CJ than me, because I want this to be a perspective ⁓ from someone that's actually on the front lines helping implement AI and banking. ⁓ what we'll dive through here is kind of, as we talked about, CJ, your thoughts on how this evolution from like, banking, digital banking to now this world of AI banking and what does that look like and how are you in Middlesex thinking about this? ⁓ But maybe before we start, quick introduction on yourself and the little background and kind of, you know, how you came to the point where now you're doing something as interesting and exciting as, you know, implementing agentic banking.

speaker-1: Sure. ⁓ I'm CJ. I'm the Senior Vice President of ⁓ Operations and Innovation at Middlesex Federal in Somerville, Massachusetts. And I started in the industry in 2009 at a community bank in Maine and was in charge of digital channels and things like that. And actually, I don't even know if you know this, Dylan, but we built an app way back in the day because We were a Fiserv bank and they were rolling out, what did they call it? What their mobile banking solution was. And we couldn't put advertising on it. So we ended up actually like kind of hijacking that and making their web enabled WAP experience into an app so we could put our marketing stuff around it. And it was kind of a super cool thing. And to see the evolution from that was, I won't say the actual date, but it was after 2000 and before 2010. ⁓ And so we were doing that back then and to see how things have evolved to this point is truly one of like the most spectacular things about being in this industry over the last several decades.

speaker-0: I mean, hitting on that point, like since you started in 2009, how has the way like customers have interacted with the bank changed and evolved over the years?

speaker-1: Yeah, that's a great question. I think we can all see it, right? We've had ⁓ gone from really a branch focused experience in person ⁓ to all kinds of machinations through that, know, cafes and side by side and assisted self-service in retail locations and things. then we got to... digital banking where so every nobody had to leave their home. So it went from the branch to your home and you will be able to do all of most of your transactions there ⁓ with potentially with support from the bank either phone support probably probably phone support. ⁓ You know, some of the nicer parts of that is like being able to do a shared user experience where you can get on and see what they're seeing. That was a big of development. I'm in the support of the online experience and then as people even evolve. think even in 2009, we were talking about everything's going to be mobile. Everything's going to be mobile. And, um, you know, it's sometimes you feel like a Connecticut Yankee and King Arthur's court and banking, right? Because it takes so long for things to really evolve. I think we talked about mobile first for probably five or 10 years before it really started to happen. with the iPhone and that. And ⁓ as the phones and the technology developed beyond sort of a texting IVR experience into a screen-based experience and mobile apps really took hold, ⁓ that experience moved from your house to wherever you were. But you still had to engage with that device. And I think the next part of this, what I really saw when I first saw ⁓ Payman was the ability to even take that one step further. And it's not just wherever you are, where you are, it's whatever situation you're in. If you need, if you can speak into your phone and say, hey, I really, I'm in the car. I have two teenagers, so I spend a lot of time in the car. I'm in the car driving somewhere. And that's usually when I'm thinking about personal finances and the family finances and what needs to move where. And if you can You know, some point, it's maybe not this point right now, but at some point you'd be able to just hit a button on your carplay and tell the bank, Hey, I need to move, you know, $5,000 into the five 29 plan to pay college bill or something like that. And it would be able to do it for you. And that's kind of the vision that I have for where we're headed, that it goes from, it goes from bank, ⁓ in bank, go driving to the bank to Maybe being able to do transactions from your card at ATM, limited transactions, getting cash, going to then your home, but still attached to your desktop computer, you know, or your laptop. And then going from your home to outside, but still attached to your phone. And then almost becoming detached from your phone to be able to really just have an agent that's doing stuff for you. And potentially even doing stuff. without you, but I'm not, I'm not ready to take that leap yet. That's a scary leap to take.

speaker-0: think we're a little bit away from that, but that's a, it's a fun thought exercise. And one thing to hit on as you were talking about the, just behavior change of as users come to expect more and what technology can do. ⁓ How did that even with, you know, how did you see that happen with digital banking when you saw like, how did customers behaviors change? once you guys rolled out the mobile apps, once did you see like a change in how they interacted with the bank? Did they stop coming to branches less? You know, this point of like picking up the phone and calling your bank or maybe talking about, you know, transferring money while you're in the car, did that now happen less? Did that ever happen? Like what did that bring about that maybe we can then see in the future for AI as well?

speaker-1: Yeah, I think there's a lot there because at the same time, people's expectations of banking have changed very rapidly over the last several years, right? Like they were expecting, you might have expected when you were going into the branch to do a transaction that it might take five to seven days to move that money from one bank to another. You had to write a check maybe from your old bank and then deposit it and it goes through ⁓ the whole, the federal reserve system and the checks. check system and that could take days. And then with ACH and other with CHECK21 and ACH and other implementations, we are able to reduce some of those timeframes down to maybe a day or two. And obviously now we have real time payments ⁓ and it's taking seconds. And I think people's expectations for the for the access they have to their funds, especially with, know, fintechs and things like Plaid and MX, their access to their funds, whether they're at your institution or another institution is, ⁓ they expect to have access to those funds fairly quickly and be able to move money faster than ⁓ they've ever moved it. And I think with that comes an expectation of that facilitation of that money movement is something that agentic AI can help support and hopefully do it safely. ⁓ That's obviously a key concern is the trust in the safety of it. But ⁓ going back to really the intention of what somebody is trying to do, I think that has has been a conversation because when you had very limited ways of doing things before even online banking, you expected that it would be taken care of, but you knew what you had to do as a user. And then with online banking, it became a little difficult. I mean, it has been kind of a wild west of, not wild west, but it's been a little bit harder for regular folks to understand how they can move money. and how they can move money safely and quickly if they need to. ⁓ One of the things we've heard in our industry is like the concept of wires. Everybody wants to wire money, but it's actually sort of, ⁓ instead of actually using wires, maybe they just wanna move it faster, right? So that becomes a part of the conversation when you start to talk about the options you have with electronic transaction, electronic funds transfer. and online banking and mobile banking, your support team needs to be able to understand the products to a point where they can have those kinds of conversations. And I think that ⁓ AI offers a lot in that area of understanding. Like we can train it to understand how the best way maybe to move money and to make those suggestions to the customer in a way that ⁓ potentially you would have to spend a lot of time and effort training a team to be basically everyone being a cash management specialist. Right? So some of the skills that you're better cash management folks bring to the table to like, ⁓ understanding how to move money, asking the customer what they want to do, how they want and giving them options in terms of how to accomplish that.

speaker-0: I think you hit on so many good things right there, which is like awesome. I think like really that last part, which is how do you think about, you know, free owns as you know, we're working together and implementing payment. Like there's a point where you start asking not like, you know, can this can, can this do that? Can this do that? But like, why can't it do this? Why can't it do this? And you keep wanting to like, feed it more to be able to do these things. And I think before I get into the intents, before I get into like even this part of upscaling the employees, like how did you come to the grips of like, okay, AI has traditionally been like put in customer service and cost cutting roles, but this is not like how you're thinking about this. It's more so like, how do we grow the business and how do we build an AI, like really a banking AI experience? Like how does this AI do the banking? What spurred that? Like, how did you see this evolving to this point and then feeling like, okay, now is the time to actually take this step and start implementing something like this or start doing something like this?

speaker-1: Yeah, I think it had to do with, you know, for everybody, I think it started off with like, chat GPT and some of the other things, right, and sort of asking it questions and getting to know how the how the thing works and how AI in general works. And then it really hasn't taken that long. I think the first for me personally, the first iteration was like, can this go through my email and delete all of that stuff that the 16,000 unread promotional emails that I've got in here. What am I spending time on that I don't need to spend a ton of time on? And it quickly turns to a lot of the things that people are spending time on and getting hung up on in banking. And then as I started to work with building some agents, because I've done a little bit of that, in developing some prompts, Then you realize, hey, you can use another agent to help you solve the problems with your first agent that you're doing. And it just gives you the answers and that. so it was like really, I think it's almost like, and I was talking to another colleague of mine who's been working very closely with you guys. She's like, it's like a light went on or like a, you know, an explosion in your head and you're like, Whoa, this is massive. And so then you just, when we saw the first demo of Payman, I was like, this is it. Like, do I make this the primary interface for how I'm going to interact with my bank? And I think that is probably fairly far ahead of the curve, but I think that might be one of the evolutionary in terms of where we're headed because this is like a very powerful tool. Obviously we have a lot of thinking about how that works and why it works and everything. you guys provided a lot of those answers to us in terms of the visibility and transparency we have on the backend in terms of how it's doing all of the things that it's doing. And that transparency and understanding how agents are working is... unlocks a lot of potential when you start to apply it to different industries and different situations. And I think, you know, some of the things that it changes all of the questions you would ask, like in a typical online banking implementation or online banking experience, you're asking a lot of questions about how the user is interacting with it and anticipating ⁓ their their use of it, what they're gonna run into, that's a problem, and then what their questions are. And with Payman, we'll be able to answer those questions, and then the best part of it is ask questions back, right? So it goes back to that establishing the goal and objective of the user, like you've said before, the intention of what they wanna do, because you don't have to start with, ⁓ And I think in banking, we've probably all been there in different situations where somebody will answer your question that you asked, the one you asked. They won't necessarily ask you, what are you trying to accomplish? Like in the perfect world, how does this work? And because that's a different answer, right? So if you tell me as, and I come from a design background, so I'm not a banking background. So you say to me, wanna make everybody in Massachusetts feel good about themselves with a billboard that we're gonna put up on the highway, right? ⁓ I can give you 10 different solutions for that problem, for that challenge. If you tell me you a ⁓ green smiley face, ⁓ you want a green smiley face on the highway, billboard on the highway, I can make a green smiley face billboard on the highway. then you're going to come back to me and say, that doesn't work. Nobody's happier. We did a study and nobody's happier. But you already came to me. I answered your question. You came to me with a specific question. I think that's in the reverse kind of what banking can do. If you have a problem, you have to ask the question that you want answered. If with AI, it It says, what do you want to do? What do you want to accomplish today? And the more people start to trust it, trust is a very, very important part of this. More people trust it to give them quality answers in terms of what do you want to accomplish? Then your answers at your ability to help that person achieve their goals is much broader than, and this is the great part of it. It always starts with a teeny tiny question. Right? need Mike, I need to, maybe I need to save money for college. Do I need to save money for college? Is a 529 playing gonna help me? I don't know, but it starts with, hmm, should I have another account for my kid? You know, and those types of things where AI could have a conversation. You, if you feel conversational, you can have that conversation with an agent and it can give you that broader view. Like, okay, why do you... What do you wanna use this account for? How can I help you achieve your goal either in the short term or the midterm or the long term? Because the other challenge with like investment questions people is like they're asking you questions for things that are 30 years from now. Now I am that 30 years from now now and I kind of wish I'd answered those questions differently back 30 years but I'm hoping that AI can bridge the gap between the... Okay, well, if you want to retire when you're 67 and you need to be saving X now and you're like 25 years old working at a ski area, present self, and you're like, I'm not going to be able to save that much money each month. know, like, I don't know. I'm hoping that it will be able to bridge the gap and help somebody with information that helps them make those small decisions into large decisions in the appropriate timeframe. with the appropriate information and really help people achieve their goals long-term.

speaker-0: That's what banking always was, right? Like it's about being, especially community banking, like it's about helping your community. And I've thought about this with AI, that light bulb moment you mentioned where it's like the more you use it, the more you start to feel like, I think it's one of, you're like, okay, I'm going to keep using this for a lot more. You'll probably ask AI things that you wouldn't even feel comfortable asking your banker, right? In person. but you will ask the AI version of the bank. And my question for you is, how does someone like Middlesex think about, well, do we want these conversations happening in ChatGPT or Claude? Or would we rather have the conversations happen with us within our own banking experience and be able to capture what our users are talking about? Because once they go to ChatGPT or Claude, those conversations aren't yours, right? Like, how do you guys think about that? because it feels like it brings something, a new dimension online to the bank where it's like, now I can bring back that relationship and understand my customer better, just like when it was in the branch, right?

speaker-1: 100%. And so going all the way back to when I got into the industry, one of the reasons I got into this industry was for ⁓ people's relationships with money. And at one of the first branches I designed in Maine, there was a quote, we did a huge quote on the wall with vinyl, and it's a quote from Sylvia Porter, is a financial writer in the early 50s. Great story if you check her out online. And one of her quotes is money, now money rarely, and I'm going to totally butcher this. So please don't anybody like, you know, text me later. Say you totally got that wrong. I know. ⁓ But like money rarely remains bits of paper and coins. It's a future it's security. can be frightening. And she kind of goes on to say it can be like, one of the most tragic things in somebody's life. can be one of the happiest things in somebody's life. And usually it's fluctuating between one or another of those things, right? At any given time, small business owners, I was a small business owner for 10 years. that, I know that experience very, very well. And you're always fluctuating between, this is the greatest thing ever. You know, and like, oh my gosh, how are we gonna make payroll next week? So, There's a lot and it's all driven by money. It's all driven by those pieces of paper and those coins. And why this is so important is... These agents can not only satisfy the mechanics of banking, but they can satisfy the emotive drivers behind their questions that people are asking and almost anticipate and help without being another human being. think that is, and you know, a lot of times banks haven't had the opportunity or haven't taken the opportunity. to fill those roles as a coach, as a personal finance coach. I know a lot of credit unions have worked on this and I kind of see financial education being slightly different than actually being, actually addressing the emotive drivers of money. Although there is some really great financial education that's starting to do that, to really talk to kids about the emotions around money and how it works. And I think that this tool ⁓ can really help people get there. Like you said, you're asking not a person, not somebody who you perceive is going to judge you, the agent. Now, whether the agent's judging you or not remains to be seen. ⁓ But at this juncture, probably not.

speaker-0: What? A hundred percent. And even the models are so good that they understand the emotions behind people. And the better the models get, the better that is. I think one other thing to hit on is like, generally with like a new shift in tech, it's like, okay, mobile banking, maybe it's like a younger generation, maybe online bank that I feel like with AI, because of the interface and because it's a chat box that you can either talk to or type to, I'd love to get your perspective because maybe I'm a bit biased, but it doesn't. seem like the blockers that tech brought before, depending on whatever cohort you think of it as, whether you're like a retiree or a new college student, it can support pretty much anyone because the UI is so intuitive. It comes down to when you interact with it and talk to it more, you learn more. How have you seen that evolve with just even using Payment?

speaker-1: Absolutely, that's part of it, right? Is the intuitive nature of it and how you can ask it questions to teach it things. So your questions are actually teaching the AI, right? And for example, ⁓ I'll just use another AI example. I was trying to do this ⁓ agent for ⁓ a work process and I'm sitting there and I know I've known of I've never taken a Python class or anything like that, but I could just, I can understand it to a point. So I just asked the AI to help me figure out how to do this. Now I would have had to go to somebody, you know, like a programmer and be like, I don't really understand, they'd be like, I don't understand what you're trying to do. But the AI didn't ask any questions, it asked questions back if it didn't understand what I was saying, but it just sort of did it and said, let's try that and let's try that. So you get to iterate a lot faster. Similarly with financial questions, you can rapid fire questions at it where you might not feel comfortable doing that with your banker. Maybe they don't have the answers or maybe you're like, well, I don't want to ask you to model this out 15,000 ways, but with the AI, you don't really care. And it is very intuitive in how to ask it to do those things. So different than trying to put it your... you know, download the data, put it in a spreadsheet, model it out, whether it's a household financial plan or a budget or something like that. Like you can really do a lot of that stuff yourself and you can say, no, we're not gonna be able to only eat out twice a month. So can you up that to three and see what it does? And it will do that. Now we didn't have enough resources to be doing that level of. stuff. And I think one of the things that kind of missed in the beginning was the personal finance managers. And I haven't ever witnessed great adoption for those. But with AI, I could see people really starting to leverage that personal financial management through the bank. And to your earlier point, Telen, that is so important to keep those conversations in the trusted source in the bank. with the people that are trying to keep your money safe in your privacy and your confidentiality, ⁓ know,

speaker-0: Yeah, I mean, you've hit on trust so many times during this. How do you think of this with AI? Like, how did you all feel comfortable to even start to pursue something like this?

speaker-1: ⁓ well, first all, I'd have to give a shout out to our leadership who, ⁓ John Wiseman, who is been supportive from day one and, just an amazing person, ⁓ in terms of really being able to see the landscape of what's available and what's coming in and, ⁓ for us to be very open-minded about what, can happen. and I think the other piece of it was how do we, you know, getting back to the back office, where's our human in the loop? you know, why is it, why would B be scared that the AI would answer a question? And I think one of the key things, and ⁓ John was mentioning this just yesterday, is that as an institution, we have to have that trust and we have to, that trust has to reside in, be extended through the AI implementation, right? And one of the things that even, And I've been at institutions where they've been like, ⁓ no, we can't do that. Somebody might go somewhere else. Well, if somebody asked, if they asked the ⁓ AI what the best CD rate is in the area, it can't just be our CDs, right? Because that's not gonna, that's disingenuous. Like this has to be an authentic experience. And that flows through the entire institution's attitude about it. I mean, it was really, I was really super excited to hear that because. I've been in institutions where we couldn't do certain marketing things or we couldn't, you know, we couldn't have certain messages out there because somebody might be aware that something is happening somewhere else. It's just not the, it's, it's not. And I realized that that's a strategy, but it feels inauthentic, right? So being authentic to our customers. And if we want to have the best rate, I mean, the way to handle that is, Hey, You know, this bank has a rate that's 10, but yours is only 10 basis points away and I can open it for you right now. You know, that's the answer to have. And so yes, have to be creative and you have to be thinking about the customer and the customer experience. again, what they're trying to accomplish is somebody going to give up. There are people that are going to go for 10 basis points more on a $1,000 CD all day long. But I'm banking on the fact that the bulk of them If our AI agent says, hey, ours is 10 basis points lower, but I can open that for you right now and move the money into it, that they're gonna say that is a convenience that's worth 10 basis points.

speaker-0: I think even like one authenticity is so important, always has been even more in an AI world. The other thing is like, you want to build that authentic experience because you want the customer to be talking to you about this and be able to trust you to be like, Hey, I'm thinking about moving my funds, not I'm going to go online and I'm just going to move my funds. And you won't know. Like, you know, we had to move our funds as a business when we raised from Brex to Wells Fargo. But all Breck saw was a transaction, not the reason behind why this is moving. And when you can figure out the reason behind why you can maybe help or understand how you can improve as a bank. Okay, well, this is why we lost this customer. And you don't want to know, you don't want to just, you know, at a certain point, lie to yourself on why you're losing customers you want to know so that you can improve. I think that's so spot on, so important.

speaker-1: And having access to each one of those conversations in the back end is again critically important for that data. We'll probably need an AI agent just to parse it out. ⁓ having that and having that audit trail of what AI did and being able to monitor the agents that we're creating. And it's an important distinction because we're not just throwing it out there. The bank is creating the agents alongside of Payman. I think that is so important that the bankers understand what they're creating and how. And I spoke earlier about my colleague, Karen Gringot, who has been working very, very closely with the implementation team. And ⁓ she does understand it and is now like brought up instances where we should use it in other places in the bank and how can we do this and how can we do that? And I think once you get it, It's not as scary as it is just throwing something out into like ⁓ Anthropic or Claude or whatever, know, chat GPT and those. ⁓ think having it be monitored by the bank, having, making sure the bank understands what's going on and putting our seal of approval on it will help people move forward with their decision to choose an AI solution agent.

speaker-0: I agree. Well, one, just fun story. I hope Karen's okay with me sharing this, but yesterday, Akash shared ⁓ her using the AI agent that we've been building together to send him some coffee money. And it was really cool because our entire team was so excited to see like, hey, look how natural and easy this is and like how fun it is to be like, ⁓ yes. you know, seven bucks for coffee in New York, which my immediate thing was, geez, coffee in New York is expensive. And we're just, but for our team, it's like, this is, know, when people might ask, why would agents send money? Why wouldn't I do it this way? It's like you hit on it so beautifully, which is this is just where things are going. This is how people will interact. It's like you mentioned before we talked, ⁓ you know, when mobile banking came out, they probably would have said, why are you doing it on your phone? I can just go do it on my desktop, but this is where people are interacting. people are going to expect more from an AI.

speaker-1: And it's about breaking down those barriers, And ease of use straight through. Again, now you don't have to necessarily click on your phone. mean, log in to go into Venmo or log into something and click on your phone four times and go through if you forgot your password or whatever. Obviously you get authenticated, but you can do these things. And I think that as we continue the evolution of this, those authentication methods will get more and more sophisticated. ⁓ you know, and hopefully we'll be staying one step ahead of the bad guys. ⁓ But those methods will become in that friction, because one of the things, you know, taking a page from our fintech partners success is trying to reduce the friction in the industry. And this really gives an opportunity for community banks to start to reduce that friction at the same level and in some cases potentially even ⁓ better than some of the fintech non-bank competitors that we face.

speaker-0: Well, I want to, this has been awesome. can talk to you for hours about this. This is so fun. I want to open up for any Q and a that maybe the audience might have that we can help answer anything that we didn't discuss or things that you maybe want us to hit more on happy to have CJ or I talk about this or else we will go on for the next hour and a half. So I think if you look at the Q and a on the bottom, right, you all can. ⁓ Ask any questions and we'll be happy to answer.

speaker-1: So while we're waiting for a question, have a question for you. What do you see as like the number one impact of a payment on a community bank?

speaker-0: Yes. I think it's, you're offering a truly differentiated product now to your customers that leverages the greatest technology ever created, quite frankly. Like ⁓ I think the impact is quite honestly, if delivered well and which it is, will be delivered well, is that people should enjoy like the banking relationship again. I open my bank account quite often, both on like business and personal and just maybe check or like, we'll do these things. But there's so many times, maybe my wife will come to me with a question and be like, you know, maybe she's doesn't want to ask me the question, but then she finally two weeks later asked me, but she should just be talking to her bank or banker about this. And I truly believe once you get this in the hands of the customers at the bank, it's going to be very hard for them to go back to banking. they. did before and it's going to be very hard for them not to have this experience because it's truly that much better. It's just like coding. can't imagine our engineers opening up like a Coda editor now. Like it's quite frankly not okay. And it's been three, really a year and a half since it's been improved. And I think banking will see the same thing. It'll be like,

speaker-1: Yeah, go down that trajectory. Interesting. We do have a couple of questions. ⁓ One of them is one of the bullet points was about compliance and audit. Can you cover that from an institution perspective? ⁓ Yeah, I mean, we are still, you know, under the regulatory framework. I mean, there's no question about that. think one of the best parts of it is that we are literally engaged, heavily engaged with writing the prompts for the agent, writing specific agents for specific uses, and doing that with a banker that has ⁓ decades of experience and compliance across the board. Reggie, UDAP, all of the regulations and things. I think The way we're framing it up is this isn't really doing anything different from a customer information standpoint. The other piece of it is some of the newer regulations around AI, human in the loop. It doesn't do things automatically unless the customer gives it that okay. And that's one of the things that we're gonna, maybe eventually, but not right now. So every time there's a transaction, there is a record of the customer saying, yes, I wanna complete this transaction. So that's an important piece. ⁓ How we're gonna audit this, that's a great question. I don't know as the regulators have fully come up with that, but one of the things that we do in our institution is try to anticipate what an audit would look like and then inform our current cadre of auditors of what we're doing, try to educate them and bring them along and say, okay, ⁓ here's what I would like you to take a look at and make sure We're doing that. And also we are very ⁓ involved with our primary regulator, the OCC, in terms of keeping them apprised of what we're doing and why we're doing in any compliance or audit issues that, or compliance issues that we may be facing. But it's all built on APIs to the core. there really are no, you know, there's an audit trail in terms of transactions and things like that. ⁓ This is what we live and breathe this every single day. And I would, if anybody has any specific questions, because I'll probably learn from the questions as well. And we're taking a very metered approach. Like this is in UAT right now. We'll probably do a friends and family and gather more information. And that's when we'll really start to put our audit and our compliance management team to the test to develop an audit plan and a compliance management program for it.

speaker-0: And I'll add just to a little bit to that is like the way the payments made is on the existing Grails and the existing digital banking provider that exists. don't, Payman or the AI agent doesn't change that. And then on top of that, as we're building this out, like you get to see the chain of thought of how the AI agents thinking and which, why it took a certain action and when the human approved. it's very purpose built to help, you know, as we work together to show why this agent made this decision. when it called a certain tool call, which APIs it hit, who asked it to do this. ⁓ sometimes you already have all the existing safety and policies and everything built into place with the core and then the digital banking provider. And then Payman comes in and says, well, we're not going to change any of that. We're also going to add our own safety layer as well. And then also our own control plane. So fun, great question. ⁓ I think here ⁓ we have, cause I think Scott, Sanjay, know you guys, Jonathan are stable coins actively being considered for this use case. CJ, I'll let you from payment perspective, no, but I'll. Let you go

speaker-1: It's on stable coins are on our radar. ⁓ They're we're taking that very ⁓ sort of stepwise as well. It's a little bit because there's so much work to do on just basically your everyday dollar transaction. I think we're trying to only bite off as much as we can chew at one time. But ⁓ stable coins. You know, not totally sure how they interact in this sort of a domestic transactional facility right now, but we are definitely stable coins are on our radar. I just don't sure how it works with payment and we're not doing any sort of international transactions at this point or ⁓ cross border work with with our agent. I mean, we're really trying to keep it fairly domestic and in dollars, but It's a great question and I think one that everybody should be thinking about.

speaker-0: I see. Yes. And then we have one minute left. I'll answer quickly Michael's question, which he asked, are you, mentioned building prompts with compliance involved. Are you referencing predetermined prompts that the customer slash member could use to speak with the agents? Are you referencing something deeper in the AI software that controls the kind of responses? Both Michael, we have our own AI engine that takes a combination of LLMs running parallel process, all of this stuff. combo with prompts and building with CJ where we will then work with the team to run through what we call intents and see if the AI agent can complete these intents to the standard that the bank would like to see. So it consists of prompts, AI engines, all of these things, but great questions all. ⁓ We will be happy to answer all of these as well, you know, over email. Please feel free to reach out to us. I really appreciate you all taking the time out of y'all's day. CJ, thank you so much for taking the time out of your day. This was fantastic. We have to do this again. And shout out also to John for also having the leadership and the vision to help us and work with us. Excited to see you guys next week. And ⁓ one last thing is we do have for anyone, ⁓ we will be sending like a little AI assessment just to... If you want to take it, it's a great way to understand like where your bank sits with how they're thinking about AI compared to other banks. So we will be sending an email after this. again, truly thank you all for your time. CJ, any closing words?

speaker-1: It's been a pleasure. Thank you. anyone would like to get in touch, yeah, I'd be happy to chat with anybody.

speaker-0: Awesome. Thank you all. Have a great day.