Automation has been part of healthcare revenue cycle management (RCM) for decades. From simple dialers to EHR integrations, each wave of technology has promised new efficiencies. But as AI agents evolve to offer real-time reasoning and adaptability, the conversation is shifting once again.

In an Office Hours episode, Evan Martin, VP of Revenue Cycle Management at ZoomCare and host of the Wilshire IT Revcast, joined Stuart Newsome, VP of RCM Insights at Infinx, to explore the human role in an increasingly automated revenue cycle. You can listen to their full conversation here.

Beyond Automation: The Need for Strategic Human Oversight

While automation has brought powerful new tools to RCM, Evan emphasized that it’s not a one-size-fits-all solution. “Just because you can automate something,” he said, “doesn’t mean you should.”

Instead, revenue cycle leaders need to treat automation like any other team member: delegated to, but closely overseen. Evan stressed the importance of training existing staff not just to perform traditional billing tasks, but to audit and monitor automated processes. Staff must be prepared to identify gaps, troubleshoot issues, and suggest technology improvements—essentially becoming the quality control layer between the machine and the mission.

Downtime Drills: A Critical (and Overlooked) Preparation

In a particularly striking moment, Evan shared a warning many RCM teams overlook: “Downtime isn’t a matter of if, but when.”

Whether due to cyberattacks, natural disasters, or unexpected system failures, Evan urged organizations to conduct “downtime drills”—simulating what happens if automation disappears. “Just like you prepare for an earthquake or tornado,” he said, “you need a plan for how your revenue cycle keeps operating without technology.”

He recommended running regular drills where technology is purposely shut down, ensuring that teams know how to switch to manual workflows when needed—before disaster strikes.

Automation’s Real Value: Amplifying Expertise, Not Replacing It

When it comes to denial management, Evan explained that technology alone won’t solve systemic problems. Instead, leaders need integrated strategies:

  • Contract Language First: Getting payer contract terms right can prevent many denials before they happen.
  • Data, But Smarter: It’s not enough to have piles of denial data; organizations need systems that identify whether denials are policy-driven or human-error-driven—and act accordingly.
  • Speak the Payers’ Language: Evan shared a clever strategy: training AI agents to “translate” clinical documentation into payer-preferred wording before submission, improving approval rates without burdening providers.

Patient Satisfaction: The Hidden RCM Metric

One unexpected insight from Evan’s conversation: RCM teams rarely have access to patient satisfaction data. Yet they are often the first and last contact patients have.

Evan called for organizations to survey patients about registration, billing, and collections experiences, and to act on that feedback. In a competitive landscape, he noted, “RCM can make or break whether a patient returns.”

Setting Realistic Expectations for AI

Finally, Evan cautioned against getting swept up in “AI hype.” Leaders need to press vendors to clearly explain what their solutions actually do—whether it’s true AI, robotic process automation, or something in between.

He also advised treating any new automation implementation with caution: start small, conduct rigorous audits, and maintain human oversight every step of the way.

Looking Ahead: A 60/40 Future

Evan predicted that within 4-5 years, the revenue cycle workforce will likely be 60% automation and 40% human expertise. But crucially, those humans will be auditors, problem-solvers, and strategic thinkers—not just process executors.

“Automation will be our frontline,” Evan said. “But the human edge will always be our safety net.”

Want more insights? Check out other Office Hours sessions on-demand here.