Healthcare revenue cycle management (RCM) sits at an inflection point. Advances in AI are expanding what is possible across healthcare, with generative models expected to drive meaningful productivity gains.
That future has not arrived for most organizations.
Administrative complexity continues to shape day-to-day operations, and RCM leaders are under pressure to deliver stronger financial performance while maintaining a consistent patient experience. Teams are working harder than ever, but effort is not the limiting factor. System design is, and this is the challenge of an in-between era.
The in-between era exposes the limits of effort
In this environment, more effort doesn’t equal better outcomes because the work itself isn’t structured for consistency, and RCM operations still depend on coordination rather than execution. Email threads start to function as workflow documentation with spreadsheets detailing system decisions, keeping critical knowledge with individuals and not in a structured, auditable place. This is a system that can’t be repeated or scaled, and performance breaks down.
For example, while treated as a volume issue, delays in accounts receivable can be traced back to how work is routed. Denials take time to correct, and yet, the conditions that created them remain in place. Teams spend too much time moving work between people and departments instead of resolving it. All that effort doesn’t produce consistent results, and it will only get worse as RCM gets more complicated. Variation will show up in places that used to feel predictable, and financial performance becomes less stable.
Adding more tools won’t solve the problem. It may even amplify it. The system itself is not designed to execute reliably.
RCM needs an operating model, not another tool
What RCM requires now is a shift in how execution is designed.
When execution still depends on coordination across people, systems, and workarounds, it will become inefficient as complexity increases. The alternative is to design execution into the system itself, where work moves through defined pathways, ownership is explicit, priorities are visible, and outcomes are measured as they happen.
When that structure is in place, performance becomes predictable because it no longer relies on individual effort.
Four layers that drive reliable execution
The next phase of RCM depends on four reinforcing capabilities.
Workflow orchestration – Work must be structured into clear queues with defined ownership and prioritization. Specialty-specific playbooks for accounts receivable, denials, prior authorization, and charge posting create consistency in how work moves through the system. Metrics shift toward execution, including touches, cycle time, resolution rates, and collection efficiency. The goal is to reserve human judgment for the work that requires it.
Agentic AI tied to outcomes – Traditional automation focuses on discrete tasks and rigid rules. Agentic AI introduces goal-based execution. Systems operate with an understanding of intent, rules, and context, allowing them to recommend or execute the next best action. With appropriate governance and auditability, this reduces stalled accounts receivable, improves denial prevention, and accelerates prior authorization workflows. It aligns automation with outcomes, not just activity.
Document and data automation that completes work – In RCM environments, the focus has been on data extraction or access. That alone does not move the revenue cycle forward. Value is created when information is classified, routed, and integrated into workflows with exception handling. The objective is to finish the work associated with that data.
Workflow intelligence that informs action – Reporting has traditionally focused on retrospective performance. What is needed now is operational visibility that informs decisions. Leaders need to see where work creates backlogs, why delays occur, which payer interactions lead to avoidable issues, and how automation performs in production. Understanding these areas supports continuous improvement and stronger financial performance.
The question leaders should be asking now
This reframes the central question for RCM leaders: The issue is no longer whether to adopt AI, but whether the organization has built a system capable of executing it reliably. Without that foundation, new technology will always struggle to deliver.The most important strategic question in RCM is no longer, “Should we adopt AI?”
That question is already behind the market. The real question is: Have we built an operating system capable of executing consistently, with or without AI?
And just behind it is a second question many organizations still avoid: Are we willing to redefine leadership for a world where scale comes from systems, not just staffing?
Without that shift, technology investments will continue to underdeliver. The tools may be modern. The outcomes will still be constrained by an operating model built on manual coordination, fragmented ownership, and workarounds.
That is the difference between experimentation and performance.
Execution is the new advantage in RCM
Organizations that focus on execution will reduce variability, improve cash flow, and build resilience against ongoing industry pressure. They will also be better positioned to integrate future innovations as healthcare continues to evolve.
In the next phase of RCM, financial performance will be defined by how well organizations design for execution.
Photo: Witthaya Prasongsin, Getty Images
Muthu Raju serves as Chief AI & Technology Officer at Knack RCM, where he leads AI strategy, platform engineering, and enterprise-scale transformation across healthcare and regulated industries. A technology entrepreneur at heart, Muthu has built and scaled multiple technology companies and platforms, translating AI and automation into measurable financial and operational impact.
This post appears through the MedCity Influencers program. Anyone can publish their perspective on business and innovation in healthcare on MedCity News through MedCity Influencers. Click here to find out how.
