By Ray Birch
ROCKLAND, Mass.—By every measure, artificial intelligence is reshaping financial services. But while much of the conversation has centered on chatbots, fraud detection and member-facing applications, one Massachusetts credit union has quietly found value in a far less glamorous corner of the business: the back office.
Rockland Federal Credit Union, the $3.7-billion institution that will become Arise Financial on Aug. 1, isn't using AI to replace employees. Instead, executives say they deployed it to eliminate repetitive work, improve compliance and create capacity for growth—all without adding staff. The results have been dramatic enough that the credit union is now expanding the technology beyond indirect auto lending into mortgages and other operational areas.
"It's important to understand our objective wasn't workforce consolidation," said David Cedrone, Rockland's chief lending officer. "It was really how do we scale our business and grow it with the use of technology."
That distinction may prove to be one of the more important AI lessons for credit unions.
Start Small, Scale Fast
Like many credit unions, Rockland's indirect auto lending operation depended on employees manually reviewing loan packages before funding and again during quality control. Staff compared documents against internal policies using checklists, spending roughly 20 minutes reviewing each file.
With approximately 2,000 indirect auto loans flowing through the department each month, the process consumed significant staff time while allowing only about 15% of files to receive quality control review.
Growth meant adding employees.
Instead, Rockland partnered with Kintera AI, whose platform converts institutional policies into automated workflows capable of reviewing loan documents against those standards.
The implementation took just 10 days.
David Cedrone
The Numbers Tell The Story
The operational improvements extended well beyond labor savings.
Per-file review time dropped from roughly 20 minutes to just two minutes. Manual sampling gave way to reviews of every loan file, increasing quality control coverage from 15% to 100% while creating a complete audit trail for every decision.
Cedrone estimates the project delivers approximately $250,000 in annual savings from this single workflow alone.
Perhaps more significant, however, was the effect on production.
Since the project began Rockland has almost tripled its indirect auto loan production. For Cedrone, that validates the strategy.
"If you imagine the FTE cost per loan that you're spending on that, and we do about 2,000 of these a month, the numbers start to really add up just in this one isolated area," he said.
AI That Fits The Business—Not The Other Way Around
One of Cedrone's strongest recommendations for other credit unions is surprisingly conservative.
Don't redesign your institution around AI.
Instead, integrate AI into existing business processes.
"I think that's an important point for some of the do's and don'ts," Cedrone said. "Not to try and reinvent your business, but just try and improve it."
Christian Klacko, CEO of Kintera AI, believes many institutions make the mistake of searching for an enterprise-wide AI strategy before taking the first step.
"I talk to a lot of CEOs of credit unions, and the first question I always hear is, 'I don't know where to begin,'" Klacko said.
His advice: identify the biggest operational pain point and solve that first.
He compares AI adoption to Tesla's autonomous driving capabilities.
"You don't go from zero to Level Five," Klacko said. "You go Level One, Level Two, Level Three. It is a co-evolution."
That incremental approach also gives business units—not IT departments—ownership of the process. At Rockland, what began as an 11-item automated checklist has expanded to 55 compliance checks, with lending staff themselves adding new rules as business needs evolve.
Beyond Lending
The success of the indirect lending project has prompted Rockland to begin automating mortgage quality control, where loan files often exceed 600 pages.
But Cedrone sees an even larger opportunity.
Rather than focusing on individual departments, he is examining workflows that span the entire enterprise—from lending and deposits to commercial operations and other back-office functions.
"We're looking at the organization as an enterprise connection where we can benefit the most broadly for the biggest lift," Cedrone said.