In an insightful interview at GAC 2025, Sarah Snell Cooke of The Credit Union Connection sat down with Mike de Vere, CEO of Zest AI, to discuss the transformative power of AI in the credit union industry. De Vere, a seasoned expert with over 15 years in the AI space, highlighted how Zest AI is revolutionizing lending decisions and fraud protection for credit unions.
De Vere emphasized that Zest AI has been perfecting the use of AI since 2009, long before it became a mainstream buzzword. Their core mission is to help credit unions make crucial lending decisions, displacing outdated systems from the 1980s that were based on 1950s mathematical models. By leveraging AI, Zest AI enables more accurate decisions, ensures fair access to credit for millions of Americans often excluded from the system, and significantly boosts efficiency. De Vere noted that credit unions can increase auto-decisioned loans from 20% to 80%, meeting members’ desire for faster loan approvals.
A fascinating innovation discussed was “Lulu,” Zest AI’s generative AI solution. Born from a 36-hour hackathon, Lulu serves as a “lending intelligence companion”. Trained on 15 years of Zest AI’s normative data and publicly available information like NCUA Call Reports, Lulu helps credit unions accelerate their decision-making processes. De Vere explained that Lulu acts as a “great collaborator,” allowing credit unions to quickly access information on policies, performance, and pricing scenarios in real-time through a simple language model query. This eliminates the need for time-consuming phone calls and paper-based information sharing.
The conversation also delved into the critical issue of fraud, a growing concern for credit unions. De Vere introduced “Zest Protect,” an AI-powered solution designed to combat sophisticated fraud schemes. Unlike the “duct tape and chewing gum” approach of current multiple inadequate solutions, Zest Protect analyzes years of data to identify complex fraud patterns that human teams might miss. De Vere shared an example of a synthetic identity fraud ring, demonstrating how Zest Protect can learn from such incidents to prevent future occurrences for other credit unions.
De Vere’s closing remarks underscored the urgency for credit unions to embrace AI. He asserted that AI will be the foundation of the future lending ecosystem, with 80% of bank CEOs already integrating AI into their lending roadmaps. He warned that credit unions not adopting AI now risk being left behind in a rapidly evolving financial landscape.
To learn more about how Zest AI is shaping the future of credit unions and to hear Mike de Vere’s full insights, be sure to watch the complete interview!
Disclosure: Transcript below is automatically generated
Sarah Cooke
Hello. Welcome. My name is Sarah snow Cook. I’m the host of the credit union connection. We’re here at GAC 2025 I’m here today with Mike de Vere, welcome. Welcome. And he is with Zest AI, so lots of techie, techie stuff, yeah, for sure, why don’t you first introduce yourself in
Mike de Vere
the company? So Mike Devere, CEOs Zest, AI, we’ve been perfecting the use of AI for over a decade and a half. So when AI wasn’t the cause to lab where everybody’s talking about AI, we’ve actually been doing it since 2009 and we help credit unions actually with their most important decision, which is, who do I get a loan to? Right? And so it displaces an old system that was created the 1980s built off of 1950s math. And listen, there’s nothing wrong with the 1950s besides, like Elvis Plato was created. I mean, there’s some good things out there, but the math was born. Yeah, your mom was born. My parents? Well, no, my parents were 40s. But nonetheless, so there is actually a better way. By leveraging AI, it ends up with a more accurate decision. You ensure that each of your members get a fair shot at credit. 10s of millions of Americans are left out of the credit system. And at the same time, it actually drives efficiency. So instead of 20% of your loans getting auto decision, you’re able to drive that up to 80% which is what members want. They don’t want to sit around waiting for a loan. And so that’s how we apply our AI technology, especially
Sarah Cooke
our younger members, and members of the future, yeah, for sure. And so we were talking earlier, you mentioned Lulu. Talk to me about Lulu. Lulu
Mike de Vere
is a corgi dog in our office. Okay, literally. So she was birthed from a hackathon in 36 hours. And lulu’s father, Benja, got to name it after his dog. So Lulu is a generative AI solution. It’s a lending intelligence companion, and what it does, since we’ve been doing this for so long, you could imagine, we’ve got 15 years of normative data on everything from Policy to Performance to pricing scenarios. We’ve trained Lulu on publicly available data, like imagine an NCUA Call Report or the OCC and fed banking reports, things of that sort. And so our point in launching Lulu this year, and you know, you’ll see here, front and center at our booth, is our point in launching is we’re trying to help credit unions with their decision making process to be faster and so generative, AI shouldn’t just be for the largest banks. When you look at all the credit unions that are our customers. It represents what point $7 trillion of assets under management, so over 600 models. So that’s a lot of data. And so our thought with Lulu is she’s the great collaborator. She brings everybody together. So with all of the information sharing, instead of just calling somebody on the phone and saying, Hey, what’s your DTI policy, or how are your charge office things, she actually can do it for you in real time with just a simple, large language model, and ask her a question, and she’ll be able to answer. That’s
Sarah Cooke
excellent, because I know I’ve had, I was on the board of my credit until I moved last year, and I’ve had people ask me, what’s your what’s your I was I started as the first associate board member, okay? And it was like, What’s your policy on that? Yeah, it was like, literally handing paper, but yeah, now we’ve got all this technology that is awesome, and everybody has been yelling AI,
Mike de Vere
but now we welcome them all to the party. Welcome. It’s our party. Now it’s actually
Sarah Cooke
like, Oh, we’re gonna look for a purpose with a for the AI. And one of those purposes, other than the generative AI, is the Well, one of the things I mentioned to somebody the other day is that a lot more credit unions are the ones getting ransomware and other ways of getting hacked or fraud. And so it’s a eyes helped the bad guys get better. Oh my gosh, yeah. And so we need defense that is better. So talk about what y’all are doing. And
Mike de Vere
so that’s was the number one thing that our customer advisory board asked in third quarter of last year, is we need help protecting ourselves against fraudsters. As you said, fraudsters from around the globe are attacking the credit union industry sometimes that don’t have the resources to properly protect themselves, and what they’ve been able to go after is the fact that there’s these fraud solutions currently available in the US. Most credit unions are using four or five. It’s like duct tape, chewing gum, a little spit, a little dirt, just to try to protect against fraud, because they’re all inadequate. And so what we built is Zest protect, which leverages AI, and so it studies years of data to understand trends and patterns. And so imagine you’re a credit union in Northern California, and there’s a fraud ring that goes out and says, they create a synthetic person. They open three credit cards at three different credit unions, have an 850 industry score, which is great, right? And then they go to that last credit and say, I want a personal loan with my 850 credit score. And they say, a $40,000 loan. They get the 40,000 max out all of their credit cards and disappear. So that’s a pattern that a human can find. Now, AI thinks like a human, right? It looks for patterns. And so these are things that your best underwriter, your best fraud team. Could maybe put together, but she’s able to look at millions and millions of interactions and patterns. And so what we are able to do is learn from that pattern such that no other credit union will face that issue in the future. And so we’re so excited right now the adoption rate of zest protect right now, it is definitely a need, and it’s only going to get better and stronger as we go.
Sarah Cooke
Yeah, and now, so you also mentioned you’ve been running around like crazy, not just here at GAC, but around DC in general. Correct? What’s What are you most looking forward to here?
Mike de Vere
So well, first off, I always love visiting with customers. So that’s my number one thing. You know, I got to sit next to my first customer, Geneva man, who’s now the CEO of ORNL, Federal Credit Union. My number three customer, Greg Young from Hawaii, USA, but now, with 600 over 600 it’s like so great to walk the hall and be able to have an opportunity to kind of engage with them. So that’s number one. Two, meeting new people. Three, it’s also engagement with DC. So if we look at things like engagement with the current administration, the House, the Senate, the Federal Reserve right now, what we’re noticing is there’s this real interest in engagement. They want to lean in with the industry and understand more about how they can support us. And so there’s a lot of that going on right now. So a lot of security checkpoints today to get through the day. Yeah, for sure.
Sarah Cooke
Yeah. Probably a protest here and there. So always give my guests final thoughts, yeah.
Mike de Vere
What do you wanna leave our audience with? So you know, the future of a lending ecosystem at the foundation is going to be AI. So 80% of banks CEOs have all said AI is in their lending roadmap. And so the big thing for credit unions in particular is whether you’re talking about Zest or someone else is really leaning in heavy on AI. And the promise that it has to protect your members against fraud, and credit unions against fraud, it has a great promise in bringing more consumers into the credit system. And so, so we like that as well. But the issue will be, is, if you aren’t in it now and a couple years, you’re going to be left behind. And so it’s eat or be eaten, is what I would say. Absolutely.
Sarah Cooke
I totally agree. Thank you so much First, thanks so much for your time. I appreciate it.