The problem

What the customer was up against.

An 11-hospital integrated health system was leaving $42M annually on the table across patient access errors, downcoded charges and unworked denials. Patient access staff handled 1.8M registrations per year with 14% containing data errors that drove downstream denials. Denials work was prioritised by dollar value, not curability — high-curability low-dollar denials aged out.

The solution

What xyner built.

Deployed xyner with four specialist agents across the revenue cycle — Patient Access Agent, Charge Capture QA Agent, Denials Triage Agent and Patient Billing Comms Agent — fully integrated with Epic, the payer-contract repository and the clearinghouse.

The outcomes

Measured impact.

$42M annualised net patient revenue recovered; patient-access first-pass-accuracy up from 86% to 98%; denials worked within 5 days up from 41% to 89%; patient-billing CSAT lifted 14 points; staff retention in RCM improved as agents absorbed routine work.

Executive summary

At a glance.

Situation

An 11-hospital integrated health system was leaving $42M annually on the table across patient access errors, downcoded charges and unworked denials. Patient access staff handled 1.8M registrations per year with 14% containing data errors that drove downstream denials. Denials work was prioritised by dollar value, not curability — high-curability low-dollar denials aged out.

Intervention

Deployed xyner with four specialist agents across the revenue cycle — Patient Access Agent, Charge Capture QA Agent, Denials Triage Agent and Patient Billing Comms Agent — fully integrated with Epic, the payer-contract repository and the clearinghouse.

Outcome

$42M annualised net patient revenue recovered; patient-access first-pass-accuracy up from 86% to 98%; denials worked within 5 days up from 41% to 89%; patient-billing CSAT lifted 14 points; staff retention in RCM improved as agents absorbed routine work.

Industry

Healthcare · Revenue cycle

A regional multi-hospital integrated health system

Scope

United States (multi-state)

Patient access, eligibility verification, charge capture QA, denials management, patient billing comms

Duration

12 weeks pilot, 9 months full rollout

From contract signature to full rollout.

Architecture

What the deployment actually looks like.

The deployment runs in the health system's HIPAA-aligned private cloud with a single-tenant data plane; PHI never leaves the perimeter and all model inference uses HIPAA-eligible endpoints in the same region. Each agent inherits the requesting clinician or RCM staffer's role scope.

Patient Access Agent

At registration time, validates demographics, insurance eligibility, and authorization requirements in real time; corrects data errors and prompts staff with the exact field to fix.

Charge Capture QA Agent

Continuously reviews documented care against captured charges; surfaces missed charges with citations to the clinical note; routes for coder review.

Denials Triage Agent

Triages denials by curability, payer, dollar value and aging; routes to the right work queue; drafts appeals for high-curability denials with full evidence.

Patient Billing Comms Agent

Communicates with patients about balances, payment plans, financial assistance eligibility; respects vulnerable-patient policies and avoids collections-style pressure.

Payer-policy RAG

Policy-aware RAG indexed on each payer's medical policy, fee schedule and appeal rules; retrieval bounded by payer, plan and date.

HIPAA-grade audit

Every PHI touchpoint is captured; every payer interaction logged; PHI redacted on any model call that doesn't require it.

Implementation timeline

How the rollout sequenced.

A 12-week pilot covered three hospitals before expansion across the full system. Clinical and revenue-cycle leadership co-designed the deployment.

Weeks 1-3

HIPAA foundations

Deploy in HIPAA private cloud; configure HIPAA-eligible model endpoints; complete BAA paperwork; deploy single-tenant data plane.

Weeks 4-5

Integration

Integrate Epic (HB, PB, Resolute, Tapestry where applicable), payer-contract repository, clearinghouse and patient-billing platform; configure RBAC from system IdP.

Weeks 6-7

Agent configuration

Configure four agents against the system's RCM playbooks; load payer policies and fee schedules into policy-aware RAG; complete first clinical-safety testing.

Weeks 8-10

Three-hospital pilot

Live in three pilot hospitals; metrics reviewed daily; weekly business review with CFO and CMIO.

Weeks 11-12

Pilot evaluation

Pilot outcomes reviewed; expansion plan agreed; staff training scaled.

Months 4-9

System-wide rollout

Hospital-by-hospital rollout with per-hospital variations; payer-by-payer coverage extended; appeals agent introduced.

Governance & controls

How the deployment is governed.

Healthcare revenue cycle carries HIPAA, billing-integrity and clinical-safety governance.

HIPAA-grade handling

PHI tagged at write-time, access-checked at read-time, redacted on any outbound call that doesn't require it; minimum-necessary enforced at retrieval.

Clinical reviewer oversight

Every charge-capture suggestion is reviewed by a certified coder before final billing; thresholds recalibrate based on coder approval rates.

Billing integrity

Agents cannot bill for undocumented services; every suggested charge cites the supporting clinical documentation.

Patient-comms guardrails

Agents never use collections-style pressure; vulnerable-patient signals (income, age, prior hardship applications) trigger financial-assistance offers.

Audit-grade trail

Every PHI touchpoint, every payer interaction, every patient communication captured for HIPAA, billing-compliance and patient-experience audits.

What other enterprises can learn

Three transferable lessons.

Three lessons for other health systems tackling revenue cycle with agentic AI.

1

Front-end accuracy compounds backwards

Most denials are caused at registration. Fixing patient access first paid downstream dividends across the entire cycle.

2

Prioritise denials by curability, not dollar value

The system's denials team had been working highest-dollar first. Switching to curability-weighted prioritisation recovered more cash with the same hours.

3

Patient comms is a quality lever, not just a cost lever

Patient-experience NPS lifted because patients felt the communication respected them. Net effect on collections was positive.

We thought we were buying revenue-cycle automation. We got staff retention, patient satisfaction and a healthier P&L. We'd do it again in half the time.
Chief Financial Officer, regional integrated health system

Reference call available through your xyner account team; the deployment is referenced in HFMA materials on AI in revenue cycle management.

Talk to a partner

Could the same outcome work in your environment?

Tell us your sector. A senior xyner partner will walk you through a tailored plan.