Why Revenue Cycle Automation Keeps Failing Hospitals

The persistent gap between automation promises and results

Hospital revenue integrity teams have cycled through wave after wave of automation platforms, each promising to eliminate denied claims and payment gaps. Yet commercial insurers' claim denial rates have roughly doubled over the past decade, according to the American Hospital Association, and hospitals continue losing tens of millions annually to underpayments they struggle to identify.

The problem isn't lack of effort or investment. It's architectural, argues Brian Sathianathan, CTO and co-founder at Iterate.ai, in an analysis first reported by HIT Consultant. Revenue cycle technology has never been designed around the actual structure of claims data—and until that changes, automation will continue underdelivering.

Why it matters

With hospital operating margins razor-thin and mid-sized facilities losing $5 to $12 million annually to recoverable denied and underpaid claims, incremental efficiency gains no longer suffice. The shift from rules-based automation to agentic AI systems represents the first viable approach that addresses the intelligence problem at the right level of the data stack—potentially changing the underlying economics rather than just accelerating existing workflows.

Rules-based systems hit a structural wall

The dominant automation paradigm in revenue cycle management has been robotic process automation and rules-based workflow tools. These systems execute defined sequences of steps when inputs match specific formats—but that's exactly where they break down in hospital environments.

Claims data is fundamentally unstructured. It lives across multiple EMRs, billing platforms, and payer portals that weren't designed to communicate. The data arrives via raw EDI files that are highly formatted but semantically complex, reflecting inconsistent coding practices, payer-specific contract terms, and constantly shifting reimbursement rules. Rules-based tools require clean, predictable inputs that hospital revenue cycle data rarely provides.

The result: high exception rates, extensive manual intervention to keep automation running, and revenue integrity teams spending time on rework rather than pattern analysis. The Healthcare Financial Management Association has consistently found that reworking a single denied claim costs $25 to over $100 depending on complexity—before accounting for claims that never get reworked at all.

Generative AI addressed the wrong problem

The wave of AI tools entering healthcare over the past two years made genuine contributions in drafting Letters of Medical Necessity, producing patient summaries, and generating appeal language faster than human teams. But from a revenue integrity standpoint, this was the easy part.

Writing better appeal letters doesn't identify why a claim was underpaid initially. It doesn't surface patterns of specific payers consistently reimbursing below contract rates for particular DRGs, or flag filing deadlines 72 hours away on high-dollar accounts. AI made revenue cycle teams more productive writers while stopping short of changing the underlying intelligence problem.

Agentic systems operate at a different level

Agentic AI systems reason over unstructured data to determine what action is needed, then execute end-to-end. The most meaningful difference is where in the data stack these systems operate.

An agentic workflow built for claims can work directly at the EDI file level, ingesting raw transaction data holistically instead of waiting for normalization into downstream systems. This means no requirement for structured inputs or deep integrations. The system can operate across multiple EMR environments simultaneously, reconciling billing records, patient data, and payer information that have never coexisted in the same space.

Implemented correctly, agentic systems can compare actual payments against contracted rates, identify coding errors before submission, track aging accounts against filing deadlines, and find underpayment patterns across payer relationships—without humans defining every rule in advance. They learn from data rather than requiring data to conform to predetermined logic.

What health systems should ask

The agentic AI category remains early enough that the gap between genuine capability and marketing language is wide. Sathianathan recommends health system leaders push past the label and ask what the system actually does with messy data: Can it operate on raw claims without clean structured inputs? Can it detect underpayment patterns rather than just execute predefined rules? Can it run alongside existing EMR infrastructure without deep integration projects?

Those questions separate genuine capability from positioning. Recent moves by major health systems to insource revenue cycle functions rather than outsource to traditional RCM vendors signal broader dissatisfaction with a model that has prioritized volume over intelligence.

These details were first reported by HIT Consultant in an analysis by Brian Sathianathan.