What You Should Know
- Healthcare revenue cycle generative AI company AKASA announced the commercial launch of its autonomous AI platform for the revenue cycle mid-cycle, expanding beyond AI-powered prebill review into fully autonomous execution for inpatient medical coding and clinical documentation integrity (CDI).
- Addresses the most complex, resource-intensive operational bottleneck in acute care: while human medical coders typically require 30 to 60 minutes per inpatient chart (with workforce shortages causing hospital accounts to sit unreviewed for three to four days post-discharge), AKASA’s platform autonomously codes inpatient encounters in under 90 seconds post-discharge, dramatically compressing accounts receivable (A/R) days.
- Validated through rigorous third-party, blinded comparative performance evaluations testing inpatient encounters representing 65% to 85% of total hospital inpatient volume, matching or exceeding expert human coders across primary compliance metrics: MS-DRG assignment, principal diagnosis selection, clinical quality capture, and present-on-admission (POA) determination.
Enterprise Scale and Footprint
The platform launch builds on rapid growth across large academic and regional health systems:
- Massive Discharged Volume: Inpatient volume processed by AKASA has grown nearly 6x year-over-year.
- National Financial Density: AKASA’s health system client base accounts for more than $180 billion in aggregate net patient revenue (NPR).
- National Discharge Footprint: The company’s models now process encounters representing roughly 1 in every 10 U.S. inpatient hospital discharges.
Technical Capabilities: Autonomous Inpatient Coding and CDI
Unlike first-generation computer-assisted coding (CAC) tools that rely on brittle keyword lookups, AKASA deploys custom-tuned generative models that operate without human intervention on complex inpatient charts:
- Sub-90-Second Chart Completion: A human medical coder takes 30 to 60 minutes to review and code an acute inpatient encounter, often after a multi-day queue backlog. AKASA’s autonomous engine completes end-to-end coding in less than 90 seconds post-discharge, accelerating billing drops and reducing A/R days.
- Complex Multi-Specialty Autonomy: Designed to fully code acute, multi-morbid inpatient stays across all clinical specialties without human touch, with outpatient facility encounter automation scheduled to follow shortly.
- Upstream CDI Integration: Extends machine reasoning directly into Clinical Documentation Integrity (CDI), identifying documentation gaps, query opportunities, and missing secondary diagnoses to ensure chart completeness before coding assignment.
- Health-System-Specific Fine-Tuning: Rather than applying a single generalized model, AKASA fine-tunes custom models on each partner health system’s historical case-mix index, localized clinical criteria, documentation patterns, and specialty nuances.
Third-Party Clinical Validation & Benchmark Data
Addressing federal and health system scrutiny regarding AI compliance, AKASA subjected the autonomous engine to blinded, third-party evaluations:
- Broad Encounter Cohort: Validated on inpatient encounter data representing 65% to 85% of total hospital inpatient volume.
- Parity with Human Coders: Blinded performance studies demonstrated that AKASA’s AI matched or exceeded the accuracy of experienced human medical coders across primary compliance benchmarks:
- Medicare Severity Diagnosis Related Group (MS-DRG) assignment.
- Principal diagnosis selection.
- Present-on-admission (POA) indicator capture.
- Quality and severity-of-illness / risk-of-mortality documentation capture.
Upstream Revenue Integrity Meets Capital Consolidation
As highlighted in Capstone Partners’ September 2026 report, spending on autonomous coding and billing expanded by 125% year-over-year to $450 million. With hospitals squeezed by thin operating margins and rising labor costs, health systems are replacing manual business process outsourcing (BPO) with verified autonomous software layers.
The launch mirrors recent enterprise plays—such as Ensemble Health Partners integrating Penelope Health’s codified payer policy engine and Oracle Health embedding point-of-care billing rules. By unifying CDI, autonomous DRG coding, and prebill denial prevention, AKASA stops billing discrepancies at the chart level before claims ever reach a clearinghouse or payer adjudication engine.
Early adoption by institutions such as Cleveland Clinic (which has deployed AKASA’s prebill review and is exploring autonomous mid-cycle solutions) and Nebraska Methodist Health System demonstrates that large provider networks view autonomous coding not merely as an administrative efficiency tool, but as a compliance-first requirement to mitigate audit clawbacks and commercial payer denials.
“The incredible demand for healthcare is finally being addressed by advancements in AI. Multiple parts of the healthcare ecosystem will need to scale up, with documentation and coding being critical components,” said Malinka Walaliyadde, CEO and Co-founder of AKASA. “For years, an autonomous mid-cycle has been a holy grail in our industry. Today, AKASA is making it real.”
