America’s Credit Union Automates 60% of Member Calls to Deepen Member Experience

An overloaded IVR, and leadership determined to get ahead of AI
Before 2021, ACU’s member contact center ran on a basic IVR menu with no chatbot and no live chat on the website. Every question – a balance check, a routing number, a password reset – went through AVP Debbie Possin’s contact center team, including members who called in for the same information every single day. The push to change didn’t come from a crisis, however. ACU’s president has a CIO background and was already looking at where AI could take the credit union; the pandemic then pushed the case further, driving up both call volume and online activity at once.

The goals were straightforward. ACU wanted to give members help 24/7, improve the member experience, and free up the contact center to focus on the calls that actually needed a person – notcut headcount.

Same headcount, a bigger remit, and calls that used to go nowhere
Ken, their Voice AI Agent, now handles 60% of member calls – a rate that’s held steady for over a year. That consistency matters as much as the number itself: it shows the automation isn’t a launch spike, it’s just how the contact center runs now.

After hours is where the impact is starkest. Before Voice AI, the contact center closed and members got a voicemail box. Today, Ken answers around the clock, and ACU has logged 120,000+ after-hours calls since Voice AI launched in 2023 – about 20% of total call volume – that would otherwise have gone to voicemail or been lost entirely. Members know it’s AI picking up after hours, and they call anyway – that volume is itself the evidence that members trust Ken to actually solve their problem, not just tolerate it as a fallback.

For Debbie’s team, the dividing line was never about what AI could handle – it was about what it should. Ken is built to carry complex, multi-step requests on its own, but ACU designed it to hand off anything emotionally weighted – hardship, a dispute, a difficult financial decision – by design, not because the technology couldn’t manage it.
The real shift is what that design choice frees up: with AI agents absorbing the majority of queries, agents aren’t rushing through routine calls to clear the queue – they have the time to actually sit with the conversations that need a person. Automating the routine hasn’t reduced the human side of ACU’s service – it’s expanded it.

As a result of this approach, the team hasn’t shrunk – it’s grown. Call-center headcount for routine volume held steady, and she’s added staff and responsibilities since the rollout, including MSRs= dedicated to online and loan applications and oversight of ACU’s ITMs.

Member adoption followed the same pattern most credit unions see – quick for some, a short adjustment for others, then simply how members expect to bank. As Debbie explained: “Once we explained what Ken could do – getting a balance without waiting 10 or 15 minutes in our phone queue – members adapted fast.”

Five years in, ACU’s advice to a credit union that hasn’t yet adopted AI is the same approach that got them here.