Like many community banks, $2.5 billion Republic Bank of Chicago, based in suburban Oak Brook, Illinois, needs to continuously add business clients in order to grow both the commercial lending and treasury management services that fuel its revenue.
That often starts with adding their deposits. That’s what market managers at each of the bank’s 20 branches are responsible for — and they are constantly looking for new leads. Mike Minetti, the bank’s senior vice president of retail sales, says he has relied on referral agencies to identify potential clients for his market managers in the past.
“It would just give me a list of leads for businesses with revenue up to $50 million,” he says, adding that each market manager was responsible for additional research that might make an initial introduction more meaningful.
That time-consuming process is something all relationship bankers face, and it’s a problem the founders of the artificial intelligence company Sympera AI set out to fix, says Sudharshan Krishnan, the company’s head of go-to-market and solution specialist. Using the Sympera platform, banks hand that task over to AI agents that comb through public data to develop more targeted leads.
Republic Bank of Chicago in the fourth quarter of 2025 launched a pilot that allowed its market managers to use the Sympera platform. Minetti says it is easy to use. For each lead Sympera presents, the market manager can give it two thumbs-up to indicate it’s a great lead, one thumbs-up to indicate a good lead and a thumbs-down to indicate a lead is problematic. There is also a notes field where the market manager can enter any additional information to help each AI agent learn and tweak its referral process.
“So, for example, we had a lead that was fed to us last week. The market manager had been calling on that particular customer for years at that point, and she knew that person didn’t want to move their accounts. So, she just put that information in there,” Minetti says.
While the pilot hasn’t resulted in any new commercial clients yet, the indications from it were very promising to Minetti. “All told, [we had] probably about 300 leads during that time period spread out between our branches,” he says. “And the amount of in-person appointments we set was between 10 and 15, which, given that time of year, was something more than what we were getting with our other provider.”
Based on that, Republic Bank of Chicago in May decided to scale its implementation of the Sympera platform. Minetti and others are now expecting more concrete results.
Ease of Integration
Because the bank didn’t know whether the tool would prove promising during the pilot, Chief Information Officer Madhu Reddy decided to integrate Sympera into Microsoft Teams only — the platform most utilized by the market managers. But with the full implementation now underway, Republic Bank is also expanding the use of Sympera to its commercial real estate, and commercial and industrial (C&I) lending teams as well. That means Sympera will soon feed information directly into the bank’s customer relationship management (CRM) system.
Reddy says that will provide the bank with opportunities beyond prospecting for new clients. “The other benefit is that we can nurture existing relationships,” he says. “How do we increase wallet share with existing [clients]?”
Republic Bank of Chicago uses Microsoft Dynamics 365 for its CRM, but Reddy believes any bank with a functional CRM can integrate Sympera. He says it’s actually more important their relationship bankers already know how to use the bank’s CRM effectively.
“The CRM gives you the metrics, insights and what’s really happening,” he says. “Mike’s team knows that. And the fact is that he knows that before and after [Sympera]. If we were not a proficient user of this [CRM], we just don’t know, ‘Will this help or not?’ Then you’re really sort of guessing.”
Because he knows Republic Bank’s CRM so well, Minetti says that once Sympera is connected to it, multiple departments will be on the same page with all the leads it produces. “And at that point, then our commercial real estate lenders and our C&I team will be able to look to see, ‘Oh, a market manager has already called them. Here are their notes. Maybe I just need to talk to them to see how we can get an additional appointment,’” he says.
Republic Bank of Chicago is both expecting and requiring the expanded use to pay off. Minetti says Sympera agreed to terms that require a lead accuracy rate of 90%. The agreement also requires that 12% of those leads result in prospects that take in-person meetings and that 10% of those prospects actually become bank customers.
The Agentic AI Risk Factor
While agentic AI is generally considered to present more risk than generative AI, Krishnan says the Sympera platform actually presents a relatively low risk for banks because no internal data is shared externally. It has two components. One is a public cloud where agents can store the publicly available information they gather.
“And then there’s a module which is inside the bank’s environment. And that module is connecting to internal bank data, internal bank strategies and internal bank content,” he says. “That is [only] receiving the information from the publicly hosted component, which is sending the market signals and the news and the public information. And then that information is combined with internally sourced information.”
But risk experts say financial institutions must still examine exactly what AI agents are doing. In a May 2026 white paper, the global research and advisory firm Gartner wrote that all entities using agentic AI should adjust governance based on how much scope and autonomy each agent has: the power to observe; advise; act with approval; or act autonomously. Reddy says Sympera’s AI agents never move beyond the advice stage, which Gartner defines as “generating recommendations or drafts with humans executing all actions.”
“We wanted this to be a copilot, not an autopilot. When you talk about AI, there’s a whole spectrum. This is not [autonomously] doing things for you,” he says.
The Deloitte Center for Financial Services in March released its own guide for managing agentic AI risk. Co-author Val Srinivas, the center’s banking and capital markets senior research leader, says even at the advice stage banks should at least be aware of a potential overreliance on what an agent is recommending. “If an agent’s recommendations look authoritative but are not transparent, employees may accept them too quickly, rely on them too heavily or misunderstand what the system is actually saying,” he says. “In a bank, that can lead to flawed decisions, missed exceptions or improper actions even when a human is still ‘in the loop.’”
Reddy says he’s confident that due diligence was taken on behalf of Republic Bank of Chicago to ensure that won’t be an issue, even with regulators. “I think we have a very thorough process where we’ve done risk assessment, where we go through what the risks are and then how we mitigated them,” he says. “I’m very confident that they will basically have no concerns about it.”
Previous