August 13, 2026 · Patient Financial Engagement

Breaking Agentic AI Self-Pay RCM

Cedar Launches Kora Platform — Agentic AI for Healthcare's Fastest-Growing Payer: Patients

Cedar just moved from a single AI billing agent to a full platform. Kora is a suite of purpose-built AI agents for patient financial resolution — autonomous inbound voice, proactive outbound voice, two-way text — all trained on 1.5 billion patient interactions and $13.6 billion in collected payments. The results already in: 40% inbound call containment, 20% outbound pickup rate (4x industry average). This is what it looks like when someone finally builds RCM infrastructure around patients instead of insurers.

📧 Every Tuesday: The 3-Minute RCM Briefing

Payer shifts, denial patterns, and deal signals that hit your AR before you hear about them — 3 minutes, every Tuesday.

Join RCM professionals. Free. Unsubscribe anytime.

Read by leaders from:

The Problem Cedar Is Solving — And Why It's Getting Worse

Hospitals built their revenue cycle infrastructure to collect from insurers. Eligibility checks, claims submission, denial management, ERA processing — every process, every software tool, every FTE on the team was optimized for a payer that sends 835s and operates on a contracts-and-adjudication model.

Patients don't work that way. They receive unexpected bills months after care. They have high-deductible plans they don't fully understand. They ignore portals. They don't answer calls from unknown numbers. And when they do engage, they need empathy, flexibility, and clarity — not a rigid IVR menu that routes them to a hold queue.

The numbers tell you how badly the current infrastructure is failing:

40%

of collectible dollars across Cedar's platform now come from patients without insurance coverage — up 54% in just three years. The OBBBA (One Big Beautiful Bill Act) is projected to trigger an additional $68.6 billion in hospital revenue impact through 2026–2027.

Most health systems have responded by adding FTEs or outsourcing patient collections to third-party agencies. Neither scales, and both carry significant cost-to-collect implications. Cedar's thesis — and Kora's design — is that you need AI purpose-built for this payer class, not general-purpose AI bolted onto a billing portal.

What Kora Platform Actually Does

Kora is not a chatbot upgrade. Cedar is explicit about that. It's a suite of specialized AI agents that orchestrate across voice and text, integrate with existing EHR and telephony infrastructure, and maintain memory of patient conversations across an entire care episode — not just a single call.

Kora Agent	What It Handles
Autonomous Inbound Voice	Answers billing calls, verifies identity, runs real-time eligibility checks, collects payments, handles balance inquiries, insurance capture, collections guidance — without replacing existing telephony
Autonomous Outbound Voice	Proactively reaches patients for missing insurance capture, payment collection, COB denial resolution, and Medicaid enrollment assistance
Two-Way Text	Conversational billing and payment in English and Spanish — balance info, insurance questions, payment processing
Advanced Reporting	Financial, workflow, and patient experience analytics beyond standard call center metrics

The differentiator Cedar keeps coming back to is memory. Kora remembers prior conversations. When a patient calls back after a previous interaction, they don't start from scratch. More importantly, Kora maintains context across an entire care journey — resolving a denied claim, for example, can require updating insurance documentation, getting the claim reprocessed, and then collecting the remaining balance over several weeks. Kora tracks each step and guides the patient forward rather than treating every touchpoint as a fresh start.

What makes this harder than it looks

A patient who is confused about their bill, a patient who can't pay right now, and a patient who just needs to access their HSA funds require entirely different responses. Kora was trained on 1.5 billion patient interactions to make that distinction in real time — and trigger the appropriate next financial action for each. That clinical and behavioral context is what separates a data-trained system from a general-purpose LLM pointed at billing FAQs.

The Results Already In the Field

Cedar isn't announcing a roadmap. Kora has been running at health systems and producing measurable outcomes. The numbers are specific enough to benchmark against your own operation:

600,000+ patient conversations handled by autonomous inbound calling
40% inbound call containment rate — patients fully resolved without escalating to a live agent
7.4% increase in patients attempting payment vs. interactions with live agents
4 out of 5 post-call satisfaction score from patients who interacted with Kora
20% outbound pickup rate — four times the industry average for patient billing outreach
21% relative payment rate increase within 7 days of outbound contact

The outbound pickup rate is the number that stands out. The industry average for outbound patient billing calls hovers around 5%. Hitting 20% means Kora's timing, messaging, and channel selection are materially better than what health system call centers are producing with human agents — not slightly better, four times better.

"Patients don't really want to talk about their healthcare financial struggle. Kora enables them to do it in a way that feels less threatening, and we've been pleasantly surprised with the results."

— Tennille Lizarraga, Chief Revenue Cycle Officer, ApolloMD

The Data Moat Cedar Has Built

This is where Cedar's position becomes difficult to replicate. They've processed over $13.6 billion in patient payments across 63 million patients. That's nearly a decade of behavioral data on how patients respond to billing outreach — by channel, by balance range, by demographics, by time of day, by prior interaction history.

A general-purpose AI vendor entering patient financial engagement has none of that. They can run a good LLM. They can integrate with your EHR. But they can't tell you that a patient with a $450 balance who received an outbound text on day 14 post-discharge and didn't open it will respond to a voice call on day 21 with a specific payment plan offer — because they haven't trained on 1.5 billion interactions that prove it.

Kora's propensity models, campaign logic, and conversation guardrails all run on that dataset. Every new interaction sharpens the system further across all clients in the network — the more health systems use it, the better it gets for every health system using it.

The OBBBA Timing Is Not Coincidental

Cedar launched Kora now for a reason. The One Big Beautiful Bill Act is projected to trigger a $68.6 billion hospital revenue impact through 2027 — primarily through Medicaid cuts and increased patient cost-sharing that shifts more financial responsibility to individuals. More patients will be uninsured or underinsured. More balances will land directly with patients. More collections workload will fall on billing teams that are already stretched.

Kora's outbound voice agent specifically handles Medicaid enrollment assistance — helping eligible patients enroll in coverage before a balance becomes a self-pay write-off. That's not a coincidence. It's a direct response to the policy environment every hospital CFO is modeling right now.

What this means for your cost-to-collect

Patient collections have always carried disproportionate cost-to-collect relative to commercial claims. If the patient payer mix is growing — and Cedar's data says it's up 54% in three years — and your infrastructure for that payer class hasn't changed, your overall cost-to-collect is getting worse even if your payor collection efficiency holds flat. That's the problem Kora is priced against.

What This Means for RCM Teams Right Now

If you're operating a patient billing call center

The 40% inbound containment rate is the number to take seriously. If Kora is handling four out of ten inbound billing calls without escalating to a human, your staffing model for inbound volume needs to be remodeled — not just when evaluating Cedar, but as a benchmark for any vendor or internal build you're considering. The question isn't whether agentic AI will displace first-pass inbound volume; it already is at Cedar's client base.

If you're managing OBBBA exposure

Cedar's Medicaid enrollment assistance capability is operationally significant. Proactively identifying patients who qualify for coverage and helping them enroll before a balance ages into self-pay territory is one of the highest-ROI interventions available in the current policy environment. This doesn't require a full Kora Platform implementation to evaluate — it's worth isolating as a targeted capability review.

If you're evaluating patient engagement vendors

The market has no shortage of patient communication platforms. The differentiator Cedar is making is data depth and multi-step memory — not just sending a text or making a call, but carrying full context across a care episode and adapting to individual patient behavior at scale. Ask any vendor in this space how their propensity models were trained, on how many patient interactions, and over what time period. That conversation will be clarifying.

Bottom line

Cedar's Kora Platform is the most substantive patient financial resolution product on the market, and the live results justify that assessment. 40% inbound containment and 4x industry average outbound pickup aren't pilot numbers — they're production metrics from 600,000+ conversations. The timing is intentional: as OBBBA shifts more financial burden to patients, health systems that don't upgrade their patient-payer infrastructure are going to see cost-to-collect deteriorate. Cedar is betting that infrastructure is Kora.