How Credit Unions Are Exploring a Cooperative Approach to AI

This transcript has been lightly edited for readability, grammar, punctuation, and clarity. Filler words, transcription artifacts, false starts, and repeated words may have been removed.

Doug English: [00:00:00] Welcome back to C.U. on the Show. I’m glad to see Fred Eisel and Mitch Rutledge back again. And today, our new star, Anthony Volpe. Did I say that right?
Anthony Volpe: That’s correct.
Doug English: You do use that E. Some people might not use that E. Yeah. I’m glad to see that you do. Well, I, I, I’m glad you’re all back. I’m glad we’re gonna be talking about AI for credit unions.
Uh, you guys are on the cutting edge, so let’s, uh, tell, tell our listeners what are you back to talk about today? What are you up to, uh, in your work with credit unions right now? Anybody goes first.
Mitch Rutledge: I’ll kick it off. So we, uh, you know, we’ve just had an amazing partnership with the VIZO Financial, um, corporate credit union over the last two plus years now, and as we’ve been working together, we’ve just talked about new ways to serve the movement.
And, uh, at Vertice, we’ve … you know, we’re all about, how do we bring the power of AI [00:01:00] to serve credit unions? And VIZO’s all about, how do we just serve the movement and bring all kinds of capabilities to, to, uh, support credit unions? And in that, one of the things we’ve been working on is bringing foundation AI, foundational AI solutions for credit unions.
We’ve been building those inside of Vertice. And as we had this conversation, we said the best way to get this out and not just be another product from, you know, c- what we are, right? We’re a vendor to the, to the movement, was to make a CUSO that can bring a new kind of foundation AI to the movement, and that was what we’re here to, you know, share more about today.
So we launched CUltivate AI, which is a CUSO that we want to be owned by the movement for the movement, um, back at GAC of this year. Um, so we’ve been working on this idea for years now at Vertice, but the right way for this to really be impactful and for the, the movement to be, to adopt it, [00:02:00] was to, you know, form a CUSO and, and we did that w- in partnership with, um, Fred and the VIZO team.
They were great supporters of it, and he’ll tell you more about kind of their journey to that, and also with Filene, uh, the Filene Research Organization. So, you know, Filene Reser- Research Institute. You know, with their kind of knowledge and their corpus of research, we just felt like that group to fi- form, uh, this new CUSO to bring foundational AI to the movement was gonna be the most powerful way to bring this efficiently and effectively to the movement.
And so we’re super excited to talk- Talk about CUltivate AI and this new CUSO that is for the movement, by the movement, and, and this group is, you know, the found- founding, uh, group, uh, bringing it. So that’s what we’ve been working on, Doug.
Doug English: I, I wanna hear from Fred and Anthony, but, uh, Mitch, for our listeners, you, you said a, a term I wanna make sure I understand, foundational AI.
Is that- Yeah … is it appropriate [00:03:00] to talk about that now?
Mitch Rutledge: A- absolutely. So that, I should’ve framed that from the beginning, so apologies for that. We wanna, when you think about CUltivate, CUltivate is the equivalent of a ChatGPT or a Claude or a Microsoft Copilot built for credit unions and specific to credit unions.
So again, it knows what a credit union is. It’s not a general AI model, but it is a broad capability model, right? So ChatGPT has to know everything about e- anything in, in the world and on the internet or whatever it might be. CUltivate needs to know everything there is to know about credit unions and the credit union movement.
So it is trained specifically on that. But it is foundation because it is meant to know all of the things and do many capabilities inside of the credit union movement. So think of it as kind of an equivalent to those kind of, you know, what we hear of those terms frontier models or foundational AI, but this is foundational AI for credit unions.
[00:04:00] So that is what we mean by what is, you know, CUltivate being a foundational AI solution.

Anothony: Um, again, I’m Anthony Volpe. I was a co-founder of Vertice AI alongside Mitch, and I’ve been in the tech startup space for many, many years in many different industries. In fact, um, relative to so many of the people we meet in this industry, I would say I’m a newbie.
I’ve only been, uh, learning about, talking about, thinking about credit unions with credit union partners for the last five or six years. Um, but in doing so, we’re drawing from, Mitch and I are both drawing from a well of, um, experiences and successes in other [00:05:00] industries, and we’re trying to replicate them here.
So, um, regarding the question at hand, Doug, we say foundational- Um, at the risk of confusing CTOs, so hearing you say it’s CEO-oriented is good. It’s foundational in that it plays the role of other foundation models such as ChatGPT, as you said, or Claude, and it is foundational to so many of the AI initiatives and things that we would like to do as credit unions.
So, um, I wouldn’t ever want it to be, um, confused as a true from scratch foundation model per se, but it is playing the role of a foundational AI model and infrastructure. And in fact, one of the biggest value propositions that we’ll get to later, I’m sure, in this conversation, Doug, is that we will be able to use and call other tools and agents from across the very, very rich [00:06:00] ecosystem of fintechs and CUSOs that exist, and that’s why we think of it as foundational infrastructure or foundational, um, you know, pl- a platform, if you will
Doug English: I have many questions, but I’ll wait till Fred goes.
Sure.
Fred Eisel: Well, and, and I will chirp in and say, number one, just to, just to highlight what Anthony just said about being a newbie and being, uh, you know, kind of into the industry, so to speak, uh, over the last five or so many years. Um, I have got to say, and I’ve said this many a times in front of Mitch and Anthony and other folks, that Vertice ai probably and most likely is the best vendor we’ve worked with at the corporate, uh, in my career, and my entire career has been spent at the corporate credit union, as has much of our management team, uh, say the same thing.
And the reason is why we love Vertice so much is they, because they’re from outside industries, we think that’s a benefit. They think outside the box. They, they move quickly. [00:07:00] They want to fail fast, but they, they, they see the importance of what things we need in this industry. But only being in the industry for five to six years, they’ve been very integrated, very involved.
They know the lingo. They know our challenges. They know where our weaknesses are at. So while they’re new, uh, as Anthony said, they are very involved. And you, if you talk to- if anybody’s talked to Anthony or Mitch, you would think they’ve been in the industry 20-some years. So, uh, they come with a great wealth of knowledge but a very, very good understanding of the movement, uh, of our industry, only being in a short period of time.
And so, uh, that’s why we love the partnership so well, uh, and so much, and we think what they’re delivering to the, the credit unions, uh, is absolutely critical, as we’ve talked about, uh, many times in the past. AI is absolutely critical to embrace. It’s not a trend. It is something that needs to be embraced.
The larger credit unions are already implementing in some aspects. Uh, I think it’s a game-changer for the mid to small credit unions, um, that believe, “Look, I don’t [00:08:00] have the resources. I don’t have the funding. I don’t have the time.” AI is the key because you don’t have the resources. It will allow you to free up, uh, people or do things that you may be doing today that you don’t really need to do and really eliminate the manual process.
So as a corporate credit union, uh, w- we are a strategic partner. We’re, we’re evolving and have evolved into a key strategic partner for our credit unions in accelerating their success. Yes, as a corporate, we still do the old-school things like payments and money management and liquidity. Uh, but what we need to be and what we are today is working with folks like Vertice.ai, uh, and their Vertice.ai tool, their growth, uh, membership growth engine, uh, which we’ve been working with for the last couple of years.
The CUltivate piece, the Foundate- CUltivate AI, foundational AI knowledge base piece, uh, we think is absolutely critical and absolutely needed for the industry. A lot of industries are already doing it. There’s no reason why our industry, the credit union industry, that’s supposed to be collaborative and cooperative, doesn’t have a [00:09:00] shared knowledge base that we all work with.
It makes complete sense and think it’s, could be a huge piece of the industry going forward.
Doug English: Yeah. So who’s the … Y- y- you, you maybe kind of teased it a bit, like, y- you mentioned mid-size and smaller size credit unions. Uh, i- is that the, uh, potential customer for, uh, CUltivate? Like, what do you, what do you see as, like, what’s the pain point the credit union’s trying to solve, uh, and what are their, you know, alternatives that they start, like, looking at and considering before they end up coming to Cu- CUltivate?
Like, um, a- again, keep it basic as we’re trying to just, from the basics, w- how do you end up saying, “Ah, this is the credit union way to solve this problem”? What’s the problem in the first place?
Anthony Volpe: Go ahead, Mitch
Mitch Rutledge: Well, the problem that we’re solving is that credit unions need to be adopting these advanced AI capabilities [00:10:00] to drive better service of members, efficiencies in the operations, right?
So that is the problem, right? There’s more challenge, there’s more pressure, um, and, and so they need ways to do it. Other industries are adopting- Mm-hmm … these modern AI solutions. So first and foremost, there is hesitancy from CEOs of credit unions of all sizes around fear of these AI solutions, and with the public LLMs, there should be some fear and concern, right?
Doug English: Mm-hmm.
Mitch Rutledge: And so we created CUltivate to really solve for what we believe are the main concerns of using the ChatGPTs and the Clauds and the Microsoft Copilots, and those are trust, can I trust the answers? Transparency, do I know what it was trained on? And data privacy. My… the- what keeps them up at night is some branch, you know, loan officer saying, “I’m gonna just upload this entire loan file to ChatGPT [00:11:00] and say, ‘Should I approve it or not?'”
Which, you know, we can all… that’s happening. We, we can hope that it’s not, and we’re hoping not happening often, but it probably has happened at some point in the future or in the past. And so CUltivate was created to solve for those three concerns of trust, transparency, and data privacy. That was the reason that we set out to do this, first and foremost.
So how do we solve for that? One is trust. So what it gets trained on is it will know credit unions. It’s not gonna answer the question of, uh, you know, “What should I get my wife for her birthday next month?” It can’t answer that question, but it understands what indirect lending is. It understands, uh, you know, the, all of the regulations that they need to…
NCUA regulations. It understands those, uh, that data, right? It will be, uh, loaded with data that it’s trained on that will be… a counsel will identify. We’re not gonna train it on the fringes of, you know, credit union regulatory interpretations from Reddit. We’re gonna [00:12:00] train on, you know, the actual regulations, so you can trust the answers.
Transparency is critical. You’ll be able to look up and see everything that it’s trained on, right? You can’t do that with the Microsoft Copilot. If you go ask Copilot, “What are you trained on?” It will not tell you. Mm. Same with Claude, same with ChatGT- ChatGPT, which i- in, in, you know, f- full honesty, right?
When the NCUA says that you need to be able to explain what your models are trained on, they are turning a blind eye at, “Well, we use Copilot, so it’s okay,” because you do not know what Copilot is trained on. You will be able to say, “What is this trained on?” Every corpus of training data will be available for you to have that, so complete transparency.
And then finally, data privacy, right? You’ll be able to… It w- it has a PII detector that will not allow you to upload PII data into the solution. Hmm. Nice. Right? So f- critical. So again, focusing on these fundamental needs of trust, transparency, and data privacy was the starting point of [00:13:00] CUltivate AI has to be able to do that.
That is what we’re solving for with this solution. So trust, transparency, data privacy is foundational. Now, there’s other great things, right? The, the, you know, I’ll let Anthony sort of take it from there. There’s other great things that we’re thinking about that we’re gonna help solve for. Um, so Anthony, I’ll let you pick it up and, and add to that.
Anthony Volpe: Yeah, I, I, I’m sure we’ll get to those things, Mitch. I, I would just summarize everything Mitch is saying as follows: results have to come from trusted places, trusted sources of knowledge and information. And as we know, some of these public LLMs, they’re learning a lot from e- enormous training corpuses.
Doug English: Mm-hmm.
Anthony Volpe: And we can’t always be sure that the answer’s correct. We often call those hallucinations, as you know. But where are those answers coming from? And so when we talk about trust and transparency, they go hand-in-hand with we are using the arbiters of truth in this industry, who are the most trusted sources [00:14:00] of the information that we as a credit union industry need to count on.
And, and the best example of this, Doug, working in the real world is with open evidence in the medical community, and specifically around diagnosing and treating various, um, patients. Instead of using ChatGPT, which is… You can imagine a lot of physicians went there very early on. The problem was the models are collecting information and providing ideas, prescriptions, thoughts, treatment prescriptions, I mean, not necessarily medicine- Yeah
from all corners of the internet, as Mitch said, right? Instead of counting on the New England Journal of Medicine and the Journals- Mm-hmm … of the American Medical Association, the arbiters of truth in the eyes of most people in that industry. And so open evidence said we’re going to narrow the focus, narrow the source of knowledge- And that is what provides trust.[00:15:00]
Doug English: Mm-hmm.
Anthony Volpe: That’s a very important part of this. So just to give a little more color to m- what Mitch was saying, counting on the arbiters of truth and including them, not only including them, but exclusively having them, is really critical to our success.
Fred Eisel: And to, and to kind of give you some real world case study use cases, Doug, is- Oh,
Doug English: thank you for
Fred Eisel: that.
Um, y- yeah, is for example, we at VIZO Financial are just getting ready to implement Copilot within our own, um, infrastructure. When you talk to credit un- the reason why, because our risk department s- is scared to death about Ch- Chat and Claude and anything that’s in and out, coming in and out. So when you talk to credit unions, most credit unions have the same challenge.
They, their risk groups just cannot get comfortable allowing their staff to use Chat and Claude for day-to-day usage. Now, you talk to a credit union, and I have just a few weeks ago, “Oh yeah, we’re using Copilot because we’re a Microsoft user.” Okay, so you [00:16:00] implemented Copilot. So everyone’s got… In the credit union, everyone’s got Copilot, can utilize…
“Oh, no, no, no, no. I, I have it and our chief risk officer has it- … but no one else has it.” Well, that’s not solving the problem, because you don’t trust it, right? Goes back to Mitch’s, there’s a trust factor. So you talk to most credit unions, so if the CEOs are listening, I guarantee you many struggle with, uh, the risk component comes with allowing Chat and Claude.
You’re, you’re probably not gonna let your MSR at the front line use it for anything. Maybe allow the CEO, the risk person, and maybe some C-suite folks. CUltivate AI is a protected closed system that your teller that you hired yesterday can use it to- to- today to access all the information they have, because you in the risk department, as a CEO, can trust the governance behind what’s the knowledge in that system.
The second thing is a credit union will say, “Doug, I, I, I don’t need CUltivate AI. I’ve already got my own knowledge base system.” Okay, and I, what I would say is we talk about AI, we talk about foundational [00:17:00] I- AI, CUltivate AI, all these fancy terms. I would say think about what the industry used to be. We used to go to chapter meetings once a month.
Wednesday at some Holiday Inn and have crappy chapter chicken- Yes, we did. … for dinner, and we’d share ideas. And, “Hey Doug, do you have a policy? I’m thinking about doing commercial lending. We’re dipping our toe there. Do you have a policy you can share? Maybe we grab lunch and talk about it.” That’s how you used to do it, right?
Uh, or I’d call you or email you. Um, today it’s the same thing. So we have large credit unions that say, “Fred, we don’t need CUltivate AI. I’ve got my own knowledge base.” Well, you have your own credit union knowledge base. That’s only s- as smart as your knowledge base. It’s as m- as smart as me, which means I will not go to a chapter meeting and talk to Doug and get your knowledge.
I, I’m not gonna share your knowledge in how, what did you use to commercial lending? What did you do if you expand your ma- did you do indirect lending? How did you do that? Who do you use? If you have your own knowledge base, that’s great, but it’s the same as sitting at your credit union not talking to anybody, [00:18:00] ’cause you’re not gonna get any smarter than what your knowledge is.
So with CUltivate AI, if folks share their policies, procedures, along with Filene research, along with getting information from the NCUA, if I go out there sitting at my desk today and I… because there’s no more chapter meetings, there’s no more sharing of information that way. It’s shared today through things like CUltivate AI, where I sit at my computer today and type in, “I want to start a commercial lending program.
What are my first steps? What do I, what’s the regulations behind it?” Filene is noted and cited. There might be letters to credit unions from the NCUA that are noted and cited. And it might have 10, 20, 50 different sample policies. Or if I just type into the agent, “Give me a sample, uh, policy for commercial lending,” and boom, it’s done.
That’s pretty powerful, ’cause my knowledge is limited. The, the power of this is sharing the knowledge within the industry that’s protected and governed, and I can get your knowledge, and Mitch’s knowledge, and Anthony’s, and 1,000 other credit union’s knowledge to build in a knowledge base. Final piece is when we presented to our [00:19:00] board this idea, they were extremely enthused, and one of them th- said, “You know, I’ve got subscriptions to a couple other entities.
I wouldn’t need that anymore.” No, you would not. No. CUltivate AI, the beauty and the power of AI, is getting this information together, collaborated, and, and, and governed, uh, by this advisory council, and shared. So you can have your teller you hired yesterday, feel free to let her have it and have access to it.
And the wealth of information you’re gonna get from it instantaneously, constantly, consistently, I guarantee you there’s gonna be two to three subscriptions for a variety of things they’re using today that you could eliminate those, save money, for the CEOs that are listening, but way, be made way more efficient, and be- because it CUltivates under one umbrella much more consistent, non-hallucinated data.
So, uh, we’re already talking to credit unions that have some pushback and some ideas of how would I use, uh, use it effectively. Um, there’s, there’s so many use cases of [00:20:00] why this is gonna be so powerful. Folks just gotta step back and think, not get lost in the AI jargon, but how did we used to share information?
Was verbal. Today, it’s gonna be shared in one closed, controlled, uh, space.
Doug English: Awesome. That is fantastic. Um, uh, that, that makes so much sense in the values of the credit union movement, right? The cooperative principles are, are the kinda-
Fred Eisel: There, there’s other industries- Yeah … that are doing this. Open evidence in the medical industry, Mayo Clinic s- governs it.
Powerful organization there, obviously. Other industries are doing it. Our message is, uh, of all industries that should share information, collaborate, and cooperate, it should be the credit union industry. Mm-hmm. And we got a lot of information to share. Uh, there’s no reason why this industry should not have its own shared knowledge base amongst ourselves to become a stronger industry.
And so- Yeah … that’s what Vertice AI has built, and it needs to be implemented. So for the,
Doug English: for the credit unions that are, uh, in, in your beta, uh, [00:21:00] uh, or maybe you’re past beta, but in, in your, your beginning stages, uh, what, what, what were their first use cases? Like, what have they done so far? What have, what have they found to be easy and useful and intuitive?
Uh, and what are they, where do they think they’re going next?
Anthony Volpe: The credit, credit unions today, um, that we’re working with are using compliance, policy, procedures, mir- mapping those together-
Doug English: Mm-hmm …
Anthony Volpe: uh, more coherently, more consistently. Uh, there’s use cases around marketing and creating copy. That’s something that Vertice has, um, rolled out as well.
Uh, and, and, and also I would say just the idea of consistency in product information is a very- Mm … big, important role that CUltivate can play. Um, but as, as Fred mentioned, borrowing knowledge, not only from what might be a first-party knowledge base, but we think of three tiers. The second party is, let me use [00:22:00] knowledge that other credit unions have, have provided.
And the third tier is, let me use knowledge that other experts in the industry have provided. Imagine TruStage giving expertise in insurance, or Filene, of course, and all of their research. Those three tiers of knowledge make it so powerful. So learning- Mm … about things like, hey, what are some of the fintechs out there that I should be looking at if I want to grow my member base?
Or, what are the community impact projects that have been especially successful, right? These are all things, because we’re so narrowly focused, we have great expertise in and can continue to provide that trust and transparency on.
Doug English: So the initial use cases, uh, are starting with compliance and operations is, is what I’m hearing, and that makes, that makes sense.
Yeah. That should be consistent- Yes … uh, on the compliance standpoint- Yes … uh, for, for federal credit unions anyways. Yes. Uh, and then, uh, internal consistency with policies and procedures makes, makes a whole [00:23:00] lot of sense.
Anthony Volpe: Yep.
Doug English: Uh, what are the engagement, um … H- how does a credit union engage with CUltivate? Is it a s-
Do they, do they need to, like, uh, join, uh, as a Vertus mem- or I’m not … I’m sorry, a, uh, VIZO, uh, member and be a part of the collective, or is it just, uh, they can just be a part of CUltivate? How does that work?
Fred Eisel: Uh, Mi- Mitch can, uh, expand on it, but you do not need to be a member of VIZO Financial at all. We are p- uh, a founding partner. Uh, but it is easy as going out to Cult- get CUltivate AI, um, website and sign up, uh, on the website- Yeah … and sign, uh, sign up for a, uh, an individual subscription or an enterprise subscription.
Uh, but you don’t have to be a f- you don’t have to be a, um, a, a member of the corporate at all at this point.
Mitch Rutledge: We wanna make it just like you would sign up for any of those large- Mm-hmm … enterprise AI solutions. You’ll be able to go and sign up individually or sign up [00:24:00] for your credit union, and the pricing model is a per user, per month kind of subscription, just like we, you know, you have for Copilot, and we wanna be, uh, cost b- competitive, uh, against all of those.
And, you know, like the movement, the more that sign on, this is a CUSO, it’s for the benefit of the movement, we can distribute those costs a- across more people. So this is the call for, as a movement, the more people that m- become a part of this, uh, the models train, the, you know, the cost to serve goes down.
Um, but it will be, you know, on par with the other models that you have out there, Copilot or, or equivalent in terms of cost per, per user, per month kind of scenario. So, and that’s the, that’s the approach, and it’s a very easy onboarding. There’s no, you know, it’s literally a go sign up, just like you’ve used Copilot or ChatGPT or pick your model of choice.
You’ll sign up. You can get started. It will learn over time. You can connect it to more systems to learn more over time, but this isn’t some big enterprise, uh, implementation. It can be easy to get [00:25:00] started. Now, some credit unions might say, “No, we wanna, you know, sign up everybody and do kind of an enterprise approach,” and that’s perfectly fine, too.
We wanna, you know, we’re, we’re trying to make this as frictionless as possible to get people starting on this journey, uh, with CUltivate to, you know, put this in the hands of everyone in their credit union. A- as Fred said, the example of we wanna have every person in every branch have access to this to make them more efficient and effective in serving members
Doug English: Yeah, I’ve been doing a series on, uh, AI in the boardroom.
Uh, and, and I’ve had, uh, uh, numerous inter- interviews that are gonna be coming out one after the next. And, uh, AI in the boardroom is, it’s being actively used right now. Uh, and the credit unions that I’ve talked to are using, uh, traditional models, uh, and giving it, uh, strategic plans and, um, the, uh, the board packets, uh, and making the board [00:26:00] packets interactive is, is an interesting idea.
Uh, and, um, just being a, a, a second sort of opinion in the room, like, what are we not thinking of? What does the data say? What is likely to be what comes next? Just makes-
Fred Eisel: Yeah …
Doug English: a ton of sense. Um, can you, uh, think about, like, uh, again, I know, I know for a fact that credit unions are doing that right now. Uh, and they tell their model to reference the NCUA guidelines or what’s appropriate for that credit union.
What is, how would you differentiate how CUltivate would work versus what, uh, I know credit unions to be doing strictly in this lens, just strictly for the board purposes with the board packet and the strategic, uh, process. How would you, how would you th- experience the differences between those two?
Anthony Volpe: Well, for me, CUltivate represents the most data-informed Resource that you could possibly have in a situation like that.
And [00:27:00] so sometimes, as we know, board meetings don’t always go according to plan. We go off-road a little bit, and we seek answers to what seem like very critical questions at a given moment in time. Those questions often get pushed out to a data scientist who might be using AI and, and other things to, to help get answers.
CUltivate should have answers at hand. Now, that requires some more integration than Mitch just described, but that’s certainly possible. But one of the bigger opportunities here, Doug, is there are AI tools and agents, as I mentioned earlier, from many of the fintech partners that make this such a re- a rich ecosystem.
CUltivate provides the power of all of them at once from a single place. And so you can imagine a board struggling to potentially figure out wh- what, what… We have Zest AI, and we have Vertice AI, and we have Jack Henry’s new agent, and we have Callahan. What are we doing?
Doug English: Right. [00:28:00] Absolutely.
Anthony Volpe: Where one place called CUltivate has an orchestration layer that allows it to outsource requests as necessary to the partners, the tools, the agents that are best positioned to answer the questions at hand.
That is a very powerful proposition for executive-level, board-level discussion, strategy, et cetera. So again, that is a huge differentiator where we’re not asking, you know, individual solutions to stay within their box- Mm-hmm … but we have this expansive availability to, you know, the ecosystem.
Mitch Rutledge: Well, and to that- And so just, let me, let me reinforce that, Doug, because I think that j- just to say that another way, that…
So while we started in this idea of trust, transparency, and data privacy as being a primary, you know, what is the problems that we’re solving for, over the, you know, the months since we’ve launched this, this orchestration layer has become the next big challenge that we’re solving for, [00:29:00] which is- You, you should unpack that
Doug English: you have six, eight- Unpack that for the listeners-
Mitch Rutledge: Yeah …
Doug English: please,
Mitch Rutledge: Mitch. Right. So, so, you know, you, you, you’re gonna have… I mean, think about it today, how many chat windows you’re going to to ask. Everybody has an agent, right? Including Vertice, including, you know, pick every fintech that’s at, at GAC had an agent that they were demonstrating.
Well, you as a CEO or whatever C-level executive has a question, now the question is which window, which chat agent do I go ask this question to, right? Well, if it’s a lending question, you know, I’m gonna go ask, you know, my lending AI tool. If it’s a insurance question, it’s, you know, True Stage. If it’s a payments, maybe it’s Valero’s AI.
What- what- pick all of them, but now I need to know. Well, the general models aren’t gonna understand what these questions are. CUltivate will be knowledgeable of those kinds of questions to say, “Okay, this- insurance question I know I need to go ask, I’m gonna send this off to TruStage to go get the answer from, right?
Or this one I’m [00:30:00] gonna go send off to, you know, whatever my, you know, mortgage provider is or my payments provider. So this idea of having a single place chat window to ask those questions. I mean, again, CUltivate at the front end looks just like ChatGPT, whatever it might be, you know. It’s, it’s, it’s a, it’s a very common user experience.
But the orchestration on the back end is what is the true differentiator of tr- understanding the question that you’ve asked as being a credit union executive to know which agent should I pass this off to to go get that answer, so.
Fred Eisel: And a, and a, and a CEO that’s sitting here listening today is thinking, “All right, let’s think of my board meeting,” and which, which was mentioned, you’re gonna have 50 different questions going all different directions, right?
Uh, indirect lending, how’s it going? What, what’s our volumes? This or that. What are the delinquencies? Well, how our delinquencies compare. So our- we’re about 100 million in assets. What are the average delinquencies for 100 million assets under this side? You know, board members will ask general questions throughout, and how do [00:31:00] we used to handle that?
Well, let me write that down. I’ll get back to you next board meeting. I’ll shoot you an email later. Or do you… Everyone has a, has a chat, has an agent. Is somebody gonna sit there in the boardroom going to be looking at all those windows, which was mentioned? Or can somebody sit there, a C-suite person sit there and help manage with the CEO and say, “I’ll go into CUltivate.
I’ll have… You, you name me the question, I got an answer. I’ll find it, and I got one warehoused, a governed, closed, protected system that’ll answer all those questions with this shared knowledge. I don’t have to close down. I don’t have to get back to you. I have it right here in my fingertips.” Because at that board meeting, you’re gonna have questions flying at you all kinds of directions, um, you know, maybe way off the beaten trail This should be able to an- this will be, depends on how crazy the question is, but if it’s specific to that area, um, CUltivate’s gonna have that answer with the knowledge bases it’s got.
And so again, the credit union’s gonna sit there and say, “Well, I have that knowledge base.” Yeah, but CUltivate’s gonna have the industry knowledge base.
Doug English: Exactly.
Fred Eisel: So in a board, to, back, back [00:32:00] to your point, board meeting setting, a board meeting asking four or five random questions, I got your answer right now.
I got the data and I got a citation exactly where I got it from and, and where… And if you want more data, I can pull the letter of credit unions 2015 whatever. I’ll pull the research from Filene. It’s cited right here. Here’s your answer. You wanna go deeper in the weeds, I got the citation. It’s gonna be listed right there.
I’ll click on it and I’ll pull the report, and I can pull it up here on the, on the big screen if you… It’s a 20, 30-page PDF. We’re not gonna read it today at the board meeting, but I can share that with the board here during the board meeting to read later. So the- whoever’s managing that at the board level, very powerful, immediate, uh, and citations with data and meat behind it that could be shared later, uh, research and information later.
So it, it becomes extremely powerful.
Anthony Volpe: Doug, would you wanna see an example of that type of use case that Fred just mentioned?
Doug English: I would love to see that. All right. Let’s- I, I re- I really love the, the, the mental model that Fred made for [00:33:00] me, anyways- Okay … of the, the, the data. The, the bigger the data, the more the, uh, the cooperative, uh, industry comes together and, and h- and has the data in CUltivate, the more accurate, the more valuable, the more predictive, uh, the more, uh, everybody wins.
Fred Eisel: I mean, if, if you think, obvious- I went earlier and talked about chapter meetings the old school way. Even recently there was Listservs, right? Right. You go on a Listserv, uh, “Hey guys, I’m doing this. What do you think?” And you’re sharing the knowledge through a Listserv, but you gotta wait for people to respond, and share, and- Same thing.
This is just the today’s version of all that stuff, um, and information sharing in an AI-controlled, way quicker, way faster, uh, way more protected environment. And, and you’re now just not picking Doug’s brain or Mitch’s brain. We’re picking your brains, but we’re also citing Filene, TruStage, NCUA, whatever other tier supporters of this are, a deep, deep [00:34:00] knowledge base that’s protected, which is, it’s very, very powerful and
Mitch Rutledge: immediate.
I am already, I’m working on the blog post titled, No More Banquet Chicken. No more, uh- Chapter Chicken. … chapter, chapter chicken. Chapter Chicken. No more chapter chicken. No more chapter chicken. I’m working on that blog post as we speak. No, it’s great. Yes, yes.
Anthony Volpe: Doug, Anthony, let’s see what you got. Let’s see it.
Anthony Volpe: No, I’m just gonna show the example, the cascading example of here’s a basic question. We have the world’s greatest data scientist, if you will, [00:35:00] on hand immediately that can process a file pretty quickly and give some insights.
On the other hand, if we’re asking something about, as Fred said, “Hey, by the way, tell me, what are other executives saying about indirect lending?” And immediately it’s going to say, “Wow, where do we have that type of knowledge?” And it actually, I’m guessing, would go out to something like, um, Filene, which did a study on indirect lending across a variety of credit union executives, and it would come back and say, “Here’s what the research is showing about the risks.”
And so, again, it’s not confined to your own knowledge base, as Fred keeps mentioning. It is wide open to the knowledge and the tools provided by this ecosystem. And that’s the ironic part of this, Doug, success of CUltivate will depend on the true collaborative nature of the industry. Mm-hmm. If people choose to participate and be part of this movement within a movement and, and, and, and work on the [00:36:00] self-determination that comes with owning, guiding, distributing their own AI infrastructure, this will be a huge success.
There are advantages that ChatGPT or Claude cannot have. They don’t have access to Fred’s private data. But Fred and his credit union, credit unions all over, can choose to share that in a very safe way among other credit unions. And now suddenly, we, this industry, has its own AI that is actually smarter, more informed, right?
Narrowly focused, of course, but narrowly focused in a way that matters. So, um, yeah. And the more that participate, as Mitch said, the, the more inexpensive it will be and the more powerful.
Mitch Rutledge: A- and the, and the example, again, we’re new to the movement, but the example we keep going back to is shared branching, right?
This idea if, if we can have this, we’re gonna share some of these capabilities to help all of us, that is a great kind of model that we wanna replicate is h- how do we, you know, it’s not an [00:37:00] exact, uh, you know, like for like, uh, example, but hopefully that is what you understand. I, I saw you nodding, Doug. You, you get it, right?
Oh, yeah. That’s what we’re trying to do here, right? I, I, I’ve
Doug English: shared branched befo- Shared branching. I have shared branched before. Yeah. Yes.
Fred Eisel: And, and this is one of those Situations where you have over 4,000 credit unions, we expect, fully expect, you know, not every credit union’s gonna sign up for CUltivate and be involved in this, but we expect most should be part of this if they’re- Mm-hmm
truly cooperative, collaborative. Um, and, you know, I just, I think talking to credit unions throughout the last number of years especially and with AI, even the large guys, whatever large is, over a billion, over five billion, over 500 million, everyone debates what’s small, what’s mid-size, what’s large. But even the larger credit unions who, uh, I’ve been to a CEO roundtable a few weeks ago, a large credit union roundtable a few weeks ago, and they admitted, you know, [00:38:00] uh, we started building out some of these…
We can’t build it. We can’t build it quick enough. The credit unions typically, I’m not saying every one of them, but most do not have the developers and the engineers on staff to do this themselves, to build it themselves. So and if you build it yourself, great, but you also have to maintain and enhance and continue to build on that piece, let alone all the other AI tools and solutions that will be coming out from us and other folks.
So to keep up with that as a credit union, it’s gonna be very difficult. Could you partner with other folks to build that? Sure. But is that your role as a credit union to be building out fintech solutions, or is your role s- growing membership, lending to your members, and having, you have a financial picture there?
Um, so at this roundtable and what we’re hearing is a lot of large guys are like, “You know what? We, we can’t build this out anymore. We’re gonna have to partner with somebody to build this out.” That makes total sense, and even the large credit unions realize we’ve got to become way more efficient. The larger guys are the ones growing.
Those are the ones that are more profitable, but they’re also see [00:39:00] that to compete and to continue to grow, they’ve got to be way more efficient, and be- becoming way more efficient is being smarter in the fintech space, and being more efficient is being part of CUltivate AI, sharing information, have that knowledge base becomes very, very efficient.
Uh, obviously for, for the large guys, for the mid to smalls, we me- uh, talked about at the open of the, the podcast here, for the mid to smalls that don’t have the resources, it’s kind of a no-brainer. You better embrace AI.
Mitch Rutledge: Mm-hmm.
Fred Eisel: Mm-hmm. Uh, and you better embrace it quickly, uh, or it’s gonna, it’s gonna move past you.
I, at GAC, we, I talked to some credit unions that I’ve used this analogy a few times. One of the credit unions said, “I think I missed, I may have missed the AI train.” And I said, “No, you didn’t miss the train. You missed the first train. There’s two more coming. You better buy a ticket to the second two.” If you don’t, th- there is a point where it’s gonna, because it’s moving fast, ’cause people have said it’s a trend, it’s, uh, uh, and you said it, Doug, people are implementing it all sorts, in the boardroom, in lending, in growth models with Vertice AI.
It’s happening, and it’s working. It’s very effective. [00:40:00] Yeah. So you haven’t missed it, but you better embrace it and get on board one of these next two trains pretty quickly. And for mid to small, this is your game changer to, to allow you to compete with the bigger or larger credit union or local bank in your area, uh, if you embrace it, become way more efficient.
CUltivate AI is, it’s foundational because that’s one key piece for you to become way smarter and have the knowledge of all the credit union industry. Smarter, smarter with your board, smarter with your C-suite, smarter with what you’re top thinking about with your, with regard to your strat plan. Uh, it’s imperative that cranes sign up for CUltivate AI, build the knowledge base out.
It’s gonna help and, and help everyone, small and large, become way more efficient
Doug English: So CUltivate becomes your single source of knowledge, right? That’s, that’s the language- Yes … in the industry. Yeah. Yes. CUltivate becomes your single source of knowledge. Initially, perhaps, uh, the use case is compliance and, uh, and procedures, and then, uh, th- there’s just an infinite [00:41:00] number of additional use cases.
It’s just a question of, I guess we’re working our way up the trust curve, right? We’re starting with the, the, the least damaging way if you get it wrong, and are … we’re, we’re becoming, uh, more- Yes … trusting, and as time goes on, at some point we actually interact with a member, and that’s when the membership starts to feel it.
I guess they, they might feel it from, from what you’re describing already with maybe more nimble reactions, uh, uh, to go to market strategies. Mm. Uh, next best solution for member service reps when they’re interacting with a member about that member’s, uh, data set. I imagine that’s something that, that could be pretty impactful.
Are you, are you, uh, are you working on that yet or is it too early?
Mitch Rutledge: I, I mean, I … Anthony, I’ll let you speak for that, but look, this is a journey. I, I just wanna be clear, right? This is … We l- officially launched it at, uh, GAC- Yeah … um, to the, to the world. Um, you know, we’ve got betas that are in the, the, uh, plan for official GA [00:42:00] availability is, um, end of Q3, beginning of Q4, so that’s when
Doug English: we’re- So we’re right here at the launch, at the beginning of this.
Mitch Rutledge: Um, but, but, but I wanna just you know, take everybody back just 24, 36 months ago, right? The first version of ChatGPT- The dark ages. Ah. Um, right. Exactly. So, so, so all of these things are on a potential roadmap, but this is a CUSO that’s gonna be, have influence by the movement and for the movement, and so, you know, as much as we know and we have ideas, we do wanna make sure that we are driving, where is the real value add for the movement?
And so, you know, Doug, we, we can talk about the literally 60 things on Anthony’s whiteboard behind his head that’s blurred out that are all the things that we have ideas to do, but we do want to, um, you know, be methodical about what we release and when we release for the movement and thinking about that.
But I think for sure the ecosystem connectivity is gonna be a really important part of this, right? How do we connect all of the agents [00:43:00] together? Mm. Um, and, and the path for that will be, what do the credit unions want? What do the fintechs w- you know, there’s work potentially to do on some of the fintech side to make it available to connect into this, and so I think that will, will drive some of the journey.
But we’re talking to some of those fintechs already, and so, you know, that, that will be the path that comes over the next kinda 12 months of, of, you know, who can be, uh, ready to connect into the, the ecosystem, right? The connectors, as we’ve seen with all of the, you know, the other modern AI solutions. We have to build out the connector, uh, platform and get other people to make their solutions ready for connectors.
Anthony Volpe: Right. Well, and I would just add, you know, to Mitch’s point about methodical, Doug, part of the transparency is we have the ability to look at the prompts, the requests coming in from every user, and we fully intend to use those for two things at least. First and foremost, we wanna use it to support our roadmap.
We wanna understand [00:44:00] the types of things people are looking for us to solve or looking for us to solve better. And so that will drive our roadmap. And because we’re narrowly focused and we’re not having to worry about everything else that the big public eyes, we’re focused on credit union use cases, so we’ll bang those out.
But the second way that we’ll, we’ll use them is we’ll use them for training. One of the most impactful things we could do to meet our vision of, of really amplifying the contribution of each employee is to help each employee, given their role, understand what others in their role across the credit union industry are doing Imagine if we could tell them, “Hey, here’s what other compliance officers are, are doing.
Have you considered these types of things?”
Doug English: Mm-hmm. Yeah.
Anthony Volpe: Here, you’re in marketing. Here are things we’ve seen other marketers do. You should consider them, right? As we learn and the system learns, we also wanna use that and propagate those learnings to others, [00:45:00] because usage isn’t always obvious. When we come to that blank screen, that prompt-
Doug English: Mm
Anthony Volpe: right? What do I do? It’s not
Doug English: multiple choice? Good.
Anthony Volpe: That’s right. And this will be a very powerful differentiator, and one of the reasons we think executives will love to use this when we have a set of examples for every employee in their role.
Fred Eisel: We just had a risk conference, uh, in mid-April, and prior pre-conference we had an info security group, InfoSec group, uh, that met pre-conference and had a few speakers and shared knowledge and some, and, uh, some kind of, some information.
Maybe not some policies, but a lot of information. And, you know, folks that can’t get there, travel there, don’t have it in their budget, you know, we could share out. We as a corporate have a ton of collateral with regard to insight education consulting that we could put out on, uh, this as well. And so as, as Anthony said, if you’re a compliance person, tell me what’s going on.
There’s live data that just came out in April from an ins- InfoSec group, [00:46:00] risk compliance people that are talking and coming up with things and ideas and concerns that could be shared in the knowledge base. So once again, your knowledge base is as good as your knowledge base. CUltivate’s A- AI’s knowledge base will continue to g- grow and evolve and be shared to anybody.
So somebody in California says, “Well, couldn’t get to the East Coast in April for that risk group,” lot of good information sharing, kind of a synopsis of what the meeting was about, share some PDFs and some presentations. Huge piece, uh, of knowledge sharing there constantly for anyone. Uh, we have our financial conference at the end of the year.
That’s gonna be CFOs, uh, accounting, investment. So anything that anyone’s doing could be shared in that, that those that can’t attend in person, uh, can at least see the collateral and the presentations and the information shared through CUltivate, which makes it just extremely powerful.
Doug English: It’s a modern chapter meeting, Fred.
You coined that one. That was a good one too. I- It’s a modern chapter meeting. I really like it.
Fred Eisel: Trying to save you from the chapter chicken. Sit at a Holiday Inn [00:47:00] on a Wednesday night at 7:30. The old chapter chicken.
Doug English: I can taste it now.
Fred Eisel: Mm. No, you really can’t. There is no, there is no taste to it.
Doug English: It’s, it’s bouncing
Fred Eisel: back.
A lot of chewing. A lot of chewing.
Doug English: Yeah. Yeah. What an incredible idea. And there’s th- th- this, this is such a strong- model aligned with the ethos of the credit union movement, I, I, I just really wanna see it get massive legs. I love this idea for the credit union movement to be able to compete with the massive data sets that the banks, uh, have.
Obviously, the mega credit unions have, uh, massive data sets, but those are not open to the rest of the movement, and the small and mid-sized credit unions Need to, need to look at joining CUltivate as quickly as possible. We need that data when it gets put together in a safe way, uh, just creates so much power and [00:48:00] capacity for knowing what’s next.
What did I not think of? What else could it be? If there is a, a danger in this, uh, cause of a- action that I’m thinking of taking, what is it? Like, the, the ability to ask those informed questions from a safe, governed, informed, in our industry model is truly, truly outstanding. I really, really love what you guys have created.
Very interesting. I think from the standpoint of my listeners, w- what, as you all know, credit unions love more than anything, is examples. And we are super early on this, and I’m glad we’re doing this now to get the word out to all the credit unions. Look up CUltivate AI, consider joining it. And I assume, tell me if I’m wrong, if you want to be a, a, a, an owner of the data set and not just a user, then that’s when you’d look at joining VIZO as a member.
Is that how I should think about the-
Fred Eisel: No …
Doug English: thought?
Fred Eisel: No? No, you, you, [00:49:00] you can, you can, uh, be part, and I’ll let Anthony kind of g- expand on it. But you, again, VIZO membership’s not part of it. VIZO as- Okay … the corporate, we are a foundational partner, so we provide some foundational funding as part of getting this CUSO set up, uh, with the financial backing.
Vertice is providing their backing with their, uh, development of the pr- uh, platform and, and managing of the platform. Filene is a foundati- foundational partner with that initial research that goes into to get kind of things kickstarted. Um, if you want to be an owner of the CUSO, this is gonna be credit union-owned CUSO.
So if you want to be an owner of the CUSO, there is an opportunity to be, uh, various levels of partnerships and ownership levels of the CUSO. That is open. Uh, as well as information coming into the knowledge base, you can, uh, as partners like TruStage and Valero and other folks like that, can also be, uh, partial owners of the CUSO as well, depending on the information they provide.
But it has nothing to do with ownership [00:50:00] of, um, uh, membership of VIZO or anything like that. And again, if a credit union just wants to take advantage of CUltivate AI, you go online, you sign up like chat, and you pay a minimal amount per, per, per seat or an enterprise license. You don’t have to own anything.
You don’t have to be a member of anywh- anything. Mm-hmm. You just go in, sign up, you’re in.
Doug English: But there is opportu- Hey, you don’t have to share data. You can start out purely as a
Fred Eisel: user. Nope. You don’t have to share data. That’s
Doug English: right.
Fred Eisel: Yeah. If your risk group is afraid of sharing data, you don’t have to share data at all.
Obviously, it becomes more powerful the more that share. Mm-hmm. But you can sign up and don’t have to share anything, and it’s protected. Uh, but there’s an opportunity to share data. There’s an opportunity to invest in the CUSO, be a part owner. There’s an opportunity to, uh, if you’re a third party, to share even more knowledge- And become part of the CUSO that, uh, uh, that way as well.
So there’s a variety of ways to get involved, uh, with that- A
Doug English: nice careful- …
Fred Eisel: but not directly with VIZO …
Doug English: series of steps that you can take when you’re ready, when your board is on, on, on board or not. Yeah. Right? You don’t ever have to climb a ladder, but you can as your comfort increases. Uh, and, uh, uh, [00:51:00] assuming, uh, again, if you haven’t put data, uh, if you put your data into the system, I assume you can’t take it out, right?
There is no way to, to pull it back out. Once it’s in the model, it’s in the model, right?
Anthony Volpe: But you can restrict it to just your own organization’s usage. So- Oh, good. Okay … um, some, some of the ways we make data available is not necessarily to, to refine models, but it’s to be available for the model to, um, you know, retrieve on, uh, on request.
So-
Doug English: Mm-hmm. Yeah I mean, I assume it’s all, it would be all anonymized data anyways. It would be. So- Yes … yeah, yeah. It would
Fred Eisel: be. Yeah. So for a credit union that has no knowledge base, this could be their knowledge base. Some larger credit unions have already created their own knowledge base with another third party.
Um, they can continue to utilize that, share their data anonymously, uh, into the CUltivate AI platform. And once again, their current knowledge base is great for their own internal knowledge. Mm-hmm. CUltivate will provide them much more expanded knowledge within the industry. Mm-hmm. [00:52:00] Um, so, uh, if they’ve invested some time and money with another vendor for their own knowledge base, fine.
Uh, but that doesn’t mean they should not be involved in CUltivate and share their data anonymously, and then sign up for CUltivate for a small fee. And once again, they can have anyone in their, uh, credit union sign up for it, have a license, and have, uh, industry knowledge outside of their own credit union-specific knowledge base.
Mitch Rutledge: Yeah. Yeah. A- as a minimum, everyone should go to getCUltivate.ai and just show interest and get on the mailing list to get updates on what’s happening here, right? This is, uh, evolving quickly. All of those are opportunities to learn more. Uh, but this, you know, w- w- we are excited about what this is gonna mean for the movement.
And, and I think whether you’re a fintech, whether you’re a C-suite at a credit union or someone somehow connected with the ecosystem, um, you should definitely be staying up to speed on what’s happening with CUltivate AI.
Doug English: Yeah, you can only onboard so many. Are you capping, uh, your onboarding? Uh, and I mean, seriously, you can only onboard [00:53:00] so many.
Yeah. Are, are you capping that? Well- And, and you determining a number? For sure
Mitch Rutledge: there will be some. There is no set number. It… Doug, you know, look, those will be good, good problems to have. Um, but we- It
Doug English: sounds good until you have it. I can assure
Mitch Rutledge: you. Yeah,
Doug English: yeah, yeah. I have some of that, and it’s not a good time.
Mitch Rutledge: We, we, we… look, we have a… we are building this that every credit union is gonna sign on, right? So we are building this to scale to support every credit union, um, that’s out there. Um, you know, will we probably have to put some, some limits on when people get, uh, onboarded? For sure. Mm-hmm. Uh, and we’re building what that onboarding plan is as we go to general availability at the end of Q3, so.
But sign up now for sure just to stay aware and get on the list
Doug English: And there’s not a whole heck of a lot of risk. Your risk is however many subscriptions you signed up for, so you really can’t- That’s right … you really can’t go very wrong. Yeah. Very, very interesting. I want us to continue this conversation with examples.
Yeah. Uh, I want you guys to do … I want this to go really well for the [00:54:00] sake of the movement. I want, I want this industry to be able to compete with the big banks, with the fintechs, and, uh, great data, uh, recursive self-improvement, right? You got it. The next level that this is going, uh, needs to happen. So we need to have a massive data set as an industry that we use cooperatively to power everybody to be more efficient, uh, in the future to be able to survive, particularly for the small and midsize credit union.
Seems to me like a mission critical. So when you guys come back again, and this is your invitation right now, I want examples. I wanna hear about,
Mitch Rutledge: uh- We’ll, we’ll bring some credit unions back. I think the next one is- Yeah … we’ll bring some of those end users back to, to tell their stories, Doug. That’s the best way for you to hear it straight from, from that group.
Doug English: Yep, that is what resonates. There’s nothing like another credit union talking to a credit union to resonate. So- Yeah … uh, awesome thing that you guys have done. I am, I’m thrilled to support it. I wanna see it win. Uh, so check a- check out CUltivate.ai, credit union listeners, and, uh, you can look for more of this in [00:55:00] the future if, uh, these gentlemen will join me once again.
Anthony Volpe: Thank you.
Doug English: Thank you. All right, everyone. Thanks, Doug. Thanks very much. Enjoyed it. Thank you.