Fintech Meetup 2026: AI-Driven Operational Efficiency and the New Frontier of Data Access

By John San Filippo

At Fintech Meetup, which took place in Las Vegas in March, Finopotamus met with several artificial intelligence (AI) pioneers to discuss how purpose-built technology and secure data access are reshaping credit union operations. Presented herein are expert insights from these sessions.

Purpose-Built Models vs. General AI

While generic large language models (LLMs) have largely dominated the public conversation, Arjun Sirrah, founder and CEO of New York City-based Titan AI, argued that the future of banking lies in specialized intellectual property. He explained that his team spent years building a proprietary “banking ontology” and neural networks designed specifically for financial logic.

Arjun Sirrah

“We gained a lot of proficiency building our own neural networks and we saw a huge opportunity to build our own AI models for banking specifically,” Sirrah said. He noted that generic models often fail to capture the nuances of bank operations. “If you’ve got a banking ontology and banking models, you can build better banking agents,” he noted, adding that these agents are designed to automate root tasks in bank operations, such as complex search and retrieval processes that would otherwise require manual effort.

The Power of the On-Prem Security Boundary

Alekh Jindal

For many credit unions, the primary barrier to AI adoption is the risk of moving sensitive member data into the cloud. Alekh Jindal, co-founder and CEO of Bellevue, Wash.-based Tursio, explained that his company addresses this by providing “in-database AI models” that operate entirely within the institution’s own security boundary. This approach allows staff to query core databases using natural language without the data ever leaving the secure environment.

Murali Mahalingam

Tursio Vice President of Go-to-Market Murali Mahalingam noted that the industry response has been immediate. “The self-service nature of the product and the fact that we don’t have to touch or move any data seems to resonate very well with credit unions,” Mahalingam said.

Tom Foster

Vice President of Product and Project Management and Data Analytics Tom Foster of $2.6 billion Corning Credit Union, based in Corning, N.Y., confirmed the strategic value of this approach, noting that his institution tracked down the Tursio team specifically to solve data privacy challenges. “This ‘bring AI to the data’ approach was a critical security and time-to-market advantage for us,” Foster said. He observed that the system is currently used for ad hoc analysis and data pulls, with plans to expand into daily operations like fraud detection.

Solving the “False Positive” Crisis in Compliance

In the realm of risk management, Richard Meng, founder of San Mateo, Calif.-based Roe AI, highlighted a significant drain on human resources: the inefficiency of traditional rule-based monitoring. He explained that many financial institutions set simple thresholds for transactions that trigger manual reviews, leading to an overwhelming number of false positives.

Richard Meng

“Human analysts waste a ton of time just doing a bunch of false positive reviews,” Meng said. He noted that Roe AI uses investigative models to filter these alerts, allowing staff to focus on genuine threats. “Our job is to cut down those false positives and leave the ones that deserve true attention to the human analyst,” he added. By automating the investigative manual reviews for know your customer (KYC) and anti-money laundering (AML), Meng said the technology also allows compliance teams to scale their efforts without linearly increasing headcount.

Harmonizing Disparate Data for Member Insight

Finally, Chris Frank, CEO of Farmington Hills, Mich.-based CUSO Shiftmate, discussed the challenge of orchestration—getting disparate systems to talk to one another to create a unified view of the member. Shiftmate focuses on an “API-first” approach to harmonize data from cores and credit reports into hyper-personalized profiles, Frank explained.

Chris Frank

“The beauty of an agent is you determine the outcome you’re looking for, and it’s finding the best way to get to that outcome,” Frank said. He described a future where a “million little agents” work behind the scenes to find the best answers for credit union staff and members alike. Frank emphasized that his team’s deep roots in the industry provided a unique advantage in building these tools. “Everyone on our team is credit union people,” he noted, explaining that this familiarity with legacy workflows allows the CUSO to design agents that navigate the specific hurdles of credit union technology.