Interview: Q2 Holdings Data and AI head Corey Gross on navigating AI challenges and opportunities in banking 

Artificial intelligence is no longer emerging — it is now deeply embedded in banking.

While slashing costs, supercharging efficiency, and automating processes at scale, it also brings critical risks around data privacy, bias, and regulatory compliance.

In this interview, Corey Gross, the VP and head of Data and AI at Q2, discusses where the technology is driving immediate value, challenges in adoption, and how AI will transform internal banking workflows over the next few years. 

RBI: When you look at banking right now, where do you think AI is starting to have a real impact?

Corey Gross: The clearest impact is in back-office operations, where value is immediate and measurable. When AI helps a bank employee resolve a dispute or reset an account in 30 seconds instead of 15 minutes, that’s not an incremental improvement, it’s a fundamental shift in capacity. And the downstream effect matters just as much: when you reduce friction for the employee handling a case, you reduce wait time for the account holder on the other end. The impact results in better resolution times, faster issue handling, and improved CSAT scores, demonstrating where AI is creating a value chain where operational efficiency and customer experience reinforce each other simultaneously. 

RBI: A lot of banks are interested in AI, but they’re also being careful. What do you think is still making them hesitate? 

Corey Gross: A few things: trust, regulatory uncertainty, cost predictability, and workflow integration. Financial institutions need to know that any AI they deploy operates within the same compliance and security controls they already depend on. They’re also navigating a regulatory environment that’s still evolving, and moving ahead of guidance carries real risk. But the concern that often gets underestimated is economics. Institutions are watching peers absorb software costs that have ballooned unexpectedly because of token-based pricing models, and that unpredictability can lead to hesitation because it’s simply difficult to budget around. And lastly, there’s the practical question of adoption because nobody wants another system layered on top of existing workflows, forcing their teams to context-switch just to get their jobs done. 

RBI: Where does Q2 Assistant fit into that? What is it really helping bank teams do? 

Corey Gross: Q2 Assistant was built specifically to address all three barriers. It operates within the same compliance and security framework, including role-based access controls and additional governance guardrails, that underpins Q2’s broader platform, so trust is built on a foundation that already exists. We work with institutions to estimate realistic interaction volumes based on their actual case loads, helping to mitigate variability of token-based models. And on workflow integration, one of the core design principles for Q2 Assistant is for the product to feel “harmonious” with the systems already in use. The idea that AI should feel organically embedded in the workflows your team already executes, not layered on top as yet another system requiring context switching. When all three barriers are removed, adoption follows. 

RBI: From the early feedback you’ve gotten, what made you feel this was solving a real problem for customers? 

Corey Gross: We built Q2 Assistant around the workflows bank staff actually execute every day, starting with the highest-frequency, highest-friction tasks drawn from real case volume data. When we could take something that took 15 minutes and reduce it to 30 seconds, the value wasn’t abstract. But what was equally telling were the use cases we didn’t plan for, including investigating payment failures, constructing a real-time 360-degree view of a customer across transactions, entitlements, and subscribed products so a relationship manager actually knows who they’re talking to before the conversation starts. Those emerged organically from how customers chose to use the tool, and they became primary development priorities.  

RBI: Q2 already covers a lot of ground in digital banking. How do you see AI showing up across more of that product set over time? 

Corey Gross: Fraud is the most immediate and natural extension because fraudsters have been using AI for phishing and account takeover attacks for years, and the response has to be equally sophisticated. Beyond that, we see AI as the great orchestrator of complex, multi-system workflows that have historically required significant manual effort such as reconciling deposit operations, resolving account issues across systems we’ve never traditionally had access to, and pulling in context from third-party platforms to give a more complete operational picture. The broader vision is giving financial institutions leverage and scale, which offers them the ability to have complex workflows orchestrated and executed on their behalf across environments that previously required manual coordination at every step.  

RBI: How do you decide where AI can be most useful next? 

Corey Gross: Does AI make something meaningfully better, faster, and cheaper? If the answer isn’t clearly yes, we don’t pursue it, because the costs of deploying AI in the wrong places are real. Applying AI to a consumer banking interaction that’s already frictionless, for example, can turn a sub-penny transaction into a dollar of inference overhead for an outcome the account holder doesn’t value. The right places are where value is measurable and immediately recognizable, and where AI genuinely understands the systems and workflows it’s operating in. An AI that doesn’t deeply understand the data structures, terminology, and operational logic of a given environment is navigating a maze without a map. We invest where those two conditions converge between measurable ROI and deep system context. 

RBI: Looking a few years ahead, what do you think will feel most different inside a bank because of AI? 

Corey Gross: The most significant shift will be the removal of the headcount ceiling as the primary constraint on what a lean team can accomplish. Community institutions are under real pressure. They have deep institutional knowledge embedded in their people, and when those people leave, velocity suffers. They can’t simply hire their way out of it; people are the largest cost on the balance sheet, and margin pressure makes that a losing strategy against larger competitors who are aggressively automating. AI changes that equation. The institutions that will look most different in a few years are those that have embedded AI into the right workflows, not as a novelty, but as genuine operational infrastructure, and as a result can handle more volume, resolve issues faster, and compete on service quality without a proportional increase in cost. Technology has always been the great equalizer in this industry. AI is the most powerful expression of that principle yet.