One Bank Cut Lending Decisions From Weeks to Hours. Here’s How.
By Joseph A. Giannone
Simple Subscribe
Subscribe Now!
Three years into the Gen AI revolution, most financial institution strategists — especially those at small and midsized banks and credit unions — continue to advocate for test-and-learn adoption, starting small and advancing cautiously.
While that’s wise counsel, it’s only half the story. As much as experimentation and incremental progress matter, AI is far likelier to fulfill its promise when deployed enterprise-wide and with strategic purpose. Pilot programs will doubtless give confidence to your compliance team and comfort to your marketing chief, but applying AI at scale can unlock value and position even small players to seize new competitive advantages.
“AI’s payoff comes when you think big and think strategically,” said Chris Harrison, Industry Executive Director for Financial Services at Oracle. Harrison has worked alongside a broad cross-section of financial institutions, across live deployments and at deep-dive sessions hosted as part of the Oracle AI Experience series.
In an interview with The Financial Brand, Harrison articulated an approach to AI adoption that prioritizes transformative change, even for smaller institutions. “An institution’s objective should be to identify the application with the potential to deliver the greatest AI leverage,” Harrison said. The organization can then structure its AI adoption with that application in view — coordinating overall activities around it and advancing to “AI maturity” as quickly as they can sustain.
Need to Know:
- Consumer credit decisions that once took two to three weeks can now be reached in four to six hours. In commercial lending, AI reviews years of financials and pulls external benchmarking data in minutes—work that previously required multiple staff members over days or weeks.
- AI agents surface policies, procedures, and regulatory guidance in seconds instead of minutes of searching. Staff resolve inquiries faster without placing customers on hold—improving satisfaction for both customers and employees while building internal trust in AI tools.
- Instead of blasting home equity promotions to everyone with a mortgage, leading banks analyze point-of-sale data, spending patterns, and lifecycle indicators to deliver targeted offers. The goal: strengthen retention before AI agents make it even easier for customers to switch institutions.
Forging a Strategy
Harrison makes clear he is not advising banks and credit unions to leapfrog essential steps — key organizational learning opportunities — en route to achieving AI maturity. But he does urge them to ensure that each phase drives toward the institution’s larger strategic objective, or at minimum that it is designed to help identify the right strategic model for an institution’s mission and markets.
For example, an early proof-of-concept initiative might target improved operational efficiency in a specific function that impacts internal processes only and avoids use of customer or external data. Similarly, an institution at an intermediate stage might scale AI tools across functions to enhance accountholder experience — which has the added benefit of building organizational trust, because it directly impacts customer-facing employees and requires cross functional hand-offs. But as important as both examples might be on their own, each initiative should ideally focus on a subset process of a defined mission-critical function.
By the time they get to what Harrison describes as the advanced level, institutions are taking a fundamentally different approach, asking where AI can help them achieve strategic objectives in ways that typically go beyond achieving expense efficiencies. These initiatives are characterized by enterprise-level governance, clearly defined KPIs, and careful monitoring of results for business impact. This work, done well, can unlock new markets or reinvent whole business models.
Ultimately, the vision is for artificial intelligence to become part of how an institution goes to market — how it competes and especially how it grows. “The key is selecting use cases that align with where you are on the maturity curve but that also push you toward the next level of your strategy.”
Where Institutions Are Seeing Results
Harrison points to customer service as one area where AI is delivering immediate, measurable improvements that can have far-reaching implications for the institution. AI agents enable frontline employees to resolve inquiries faster by surfacing policies, procedures, and regulatory guidance in seconds, rather than requiring many minutes of searching. Instead of placing customers on hold, staff members can access accurate information immediately, resulting in quicker resolution and higher satisfaction for both customers and employees.
Lending and underwriting are undergoing a similar transformation. For consumer loans, AI ingests and analyzes tax returns and other required documents, automating completeness checks and early risk-flagging. This allows human underwriters to focus on tasks that require judgment. “If everything checks out, it moves forward,” Harrison said. “If not, that’s where the human in the loop gets involved.” Credit decisions that once took two to three weeks can now be reached in four to six hours, improving the customer experience.
In commercial lending, AI is collapsing timelines that once stretched across weeks or months. For a hotel acquisition loan, for example, AI can review multiple years of historical and pro forma financial and tax statements, while simultaneously pulling external data — such as benchmarking reports showing per-room revenue capacity, or occupancy rates versus regional averages. Complex and detail-intensive, such tasks typically require multiple staff members working sequentially across days and weeks; AI can accomplish them in minutes or hours.
Such applications dramatically compress cycle times in core business lines, and mission-critical processes. Invoking the industry adage “time kills all deals,” Harrison said: “At each step of underwriting, the quicker I can respond to the customer, the more likely they’re going to move forward with me as opposed to shopping elsewhere.”
Seizing Your Competitive Edge
For heavily consumer-focused institutions, customer engagement might prove the transformational AI use case. Specifically, institutions are applying AI to accelerate their advance from segmentation to personalization — a Holy Grail for many financial services marketers today. Rather than sending home equity loan promotions to everyone with a first mortgage, banks are analyzing point-of-sale data, spending patterns, and life-cycle indicators to deliver highly targeted offers.
“They’re using more external data, semantic analysis, to figure out what’s in the customer lifecycle that they should be more specific about,” Harrison said. The objective is to better serve accountholders and members and strengthen retention in an environment where AI agents will likely make it easier for customers to switch institutions.
And while this may reach beyond where most commercial and community financial institutions are focused today, some in the industry are looking to Formula One racing to find their competitive advantage.
Harrison described how one major asset manager is using methodology similar to Oracle Red Bull Racing’s F1 simulator — which runs a race 9,000 times before competition day — to optimize trading strategies. “They’re leveraging the same technology and the same methodology,” Harrison said, to analyze trades and determine optimal timing for moving capital. The goal is to maximize returns by running massive scenario analyses against hypothetical market conditions to help optimize timing and execution — much as F1 teams test tire compounds, fuel loads, and pit strategies to find split-second advantages. A second institution on the West Coast is now exploring similar applications, Harrison said.
For most banks, the first wave of AI adoption has delivered meaningful, but limited, gains. The banks that will pull ahead treat AI as an enterprise capability rather than a collection of tools — as Harrison puts it: “Look at what you’re trying to accomplish and how you can leverage AI to achieve your goals.” That requires executive-level governance, prioritizing initiatives that move the needle on revenue and retention, and a commitment to measuring results against business goals, from customer engagement to growth that helps banks pull away from the pack.
