MAXX 2026: Socratix AI Brings Outside Methods to Ease CU Fraud Madness
By Steve Jensen
The GoWest Credit Union Association’s six-state region – Arizona, Colorado, Idaho, Oregon, Washington, and Wyoming – gathered in Denver Oct. 6-9 for the GoWest MAXX 2026 annual conference. Finopotamus sat down with Riya Jagetia, co-founder and CEO of San Francisco-based Socratix AI, at the event.
Fraud is running rampant in the credit union industry, which continue, to fight the fraud battle using the latest technology – whether custom tools, proprietary point solutions, or whatever they can get to work.
But instead of only waging the war in a vacuum, some progressive-thinking credit unions are asking: “What’s been working against fraud in other industries? What are they doing in Silicon Valley?”
MIT-educated computer scientist Riya Jagetia had her lightbulb moment at a fraud and compliance startup. She realized AI agents could take on the investigations that were overwhelming its customers, built the company’s first AI agent for investigations, and carried that work into her next role at DoorDash. Then, sensing a narrow window to build such a company, she searched for industries in need and found credit unions and community banking.
“Our goal at Socratix,” Jagetia said, “is to build the brain for an institution. Today every platform is pitching its own AI, and each one only knows what’s in its own system. We want one shared layer of context that every AI agent and every person at the institution pulls from and contributes back to, so the work gets smarter over time and we can take on work that previously only people could do.”
That brain draws on every signal an institution generates – core banking activity, device intelligence, check images, screening hits, and analysts’ notes. “It starts with fraud and BSA/AML (Bank Secrecy Act and anti-money laundering) investigations and extends into deposit operations, loan operations, and the rest of the back office,” she said.
Finopotamus sat down with Jagetia for a Q&A session.
Q. What was the inspiration for starting Socratix AI? Was there a specific pain point in the market that you saw needed to be filled?
A: I’ve built this twice before. At my previous company, a fraud and compliance startup, I was building machine learning models and real-time rules for fraud and compliance teams. Then at a hackathon in 2023, I started playing around with GPT-3.5 (a family of large language models developed by OpenAI) and realized AI agents could take on the investigations overwhelming our customers. I went to my CEO, said “I want to build this,” and built the company’s first AI agent for investigations. I carried that work into DoorDash, pitching leadership on a team to build AI agents for fraud investigations.
Along the way, I realized this was a huge opportunity, but the window to build a company around it wasn’t going to stay open forever. So, I left to start Socratix.
When we started looking at where the need was greatest, we kept coming back to credit unions and community banks. They are dealing with the same fraud and compliance pressure as everyone else, with far smaller teams and, often, older systems. We realized the only way to do this well was to partner closely with them, so that’s where we’ve focused.
Q: How many customers do you have? And what is the biggest challenge and opportunity for you at this point?
A: We now have 15 customers, and in a short amount of time, relatively – in about one year. I think one of the greatest challenges of every vendor is when they try to bolt AI on top of a system and sell that. The big pain point with that is: Are you really going to have 15 different AI agents that only work in their own systems, running in your institution? Sure, they’re all doing meaningful work, but shouldn’t they be talking to each other and have a shared brain or context that they're able to work off? That’s the challenge. We’re talking to people about this – people who are continuing to turn on individual agents for all of their individual point solutions.
One of the exciting opportunities then is that, until now, technology just hasn’t existed that made it possible to do a lot of the work that we can do now. Like digging through documents, reviewing check images, writing narratives, writing emails to send to your members to get information from them. That's one of the things we’re building right now. We're building with new tools that are advancing every day, that are unlocking new capabilities that weren't possible before. Now is the best time to be building.
Q: Credit unions obviously have their own perceptions about technology and AI, and are used to doing things the way they do them. How would you measure their appetite for this kind of innovative AI like you’re offering?
A: It's a great question. I think we deal with that challenge by starting small and just painting the vision for them, so they know where we're trying to go, and not starting with everything all at once. So, we’ll say, “OK, let's talk about check fraud” or “Let's talk about loan document quality control – just one little pain point – and show you how AI makes it so much better.” Now we can talk about building the brain and figuring out what the long-term roadmap is for the institution to become AI native. So, I think it's walking with people rather than running ahead of them.
Q: With Socratix AI, how did you get started with that first client, then the second, to the point that you now have 15?
A: Honestly, I think the first couple of people really took a bet on us. We said, “Hey, we think we can solve check fraud a lot better than it's been solved before.” So, we did a proof of concept and made it easy for them. Obviously, they still had to spend time with us and spend some money with us, but it wasn't like we were asking them to go into a year-long contract without showing them what we could do on their data.
So, in our case, we got started through pilots, proof of concepts, and painting the vision to find those early adopters. And yes, you do have to talk to a lot of people to find those early adopters, but if you keep at it, there are usually a few who are willing to be first. There were for us.
Q: It sounds like Socratix AI is not a point solution but sits on top of their core. Does that mean that you have a certain path for their tech stack? And are you trying to partner with some of the bigger players because that's your avenue in, or not?
We haven't gotten there yet, but we have started looking into integration paths. Our first product was a check fraud solution. So, we are looking into how we can distribute that more – because we’re confident we have the best check fraud solution in the marketplace. And yes, there are a lot of them out there, but in my opinion, a lot of them are rules-based or machine-learning-based.
We believe ours is the first AI-native solution where we can store context from people who are reviewing what the agent is saying – and then tell us how to improve it over time. And so we’re trying to look for distribution channels there. Otherwise, with other workflows, we're a little bit too early. But we don't want that to stall us working with people.
Q: Who or what is your biggest competition right now?
A: I don’t think there’s really a clear competitor. I think the traditional RPA (robotic process automation) tools offer some similar capabilities and are trying to become more agentic. Still, I don't know if they’re able to build the kind of “brain” that I'm talking about – simply because when you’re that big, especially public companies, it’s more of a challenge. In Silicon Valley we say: “A freight ship is a lot harder to move than a sailboat.”
Then obviously there are individual point solution agents. And if we're trying to start with one use case, those can pose a challenge, because unless we talk about the bigger vision and get everyone bought in, it’s not an easy sell.
How are you funded? And how are you getting the word out?
Last year we raised a little over $4 million from institutional investors. But we haven’t done much otherwise in terms of publicity, especially given our great products. We don't want people to catch up with us. We want to keep our lead.
We’re also learning that the credit union community, and the community bank community, is very in-person. So we’re putting a lot into events and going to a lot of conferences. We’ve been to three events in just the past three weeks. And we’ll be going to more in the coming weeks.
How is this GoWest MAXX conference, and others like it, helping you find the right customers, and what are you learning?
At this point, we have already been to quite a few credit union events, including America’s Credit Unions’ Compliance & Risk Conference last month in Scottsdale (Ariz.). We’re now at the point that we’re meeting some of the same people over and over, telling them, “This is how things are progressing,” but we understand these decisions take a while. We're putting ourselves in their shoes, and yes, it all sounds exciting, but they’re thinking, “Can I trust these guys for the next five years? Can I trust them with our 2027 roadmap?” So it's really just meeting with people, building trust with them, telling them how the company's growing and the new customers we're adding.
I think people want to talk to other people in their region to see whether they trust a vendor, what results they've gotten, what the ROI has been, and things like that. And so whenever there are regional events like this, we try to be there because I think we want to publicize some of our marquee customers in every region – so that we can slowly grow our presence there. The hope is to get others to talk about you: “This is a real company. They’re the founders. We think they have a good vision, and you should consider working with them.” Meeting them in person, then, makes a difference. You just can’t get that from Zoom calls.
