Chris Wyatt
Venture Concept · Healthcare Adjudication
Essay

Apex Adjudication

Healthcare adjudication is buried under manual review and opaque payer policies. Apex runs an autonomous denial-resolution engine on the underlying claim data. Deterministic where the rules are clear, escalating only where they are not.

$172B
US RCM market
Grand View, 2024
$25B
Annual claims-adjudication burden
Premier, 2022
10.1%
Market CAGR, 2025-2030
Grand View

01 · The problem

The process is broken before the AI gets near it

Human coders parse dense unstructured EHR notes against hundreds of opaque payer policies. The result is a denial rate that has climbed every year since 2020 and a workflow that loses revenue on every appeal that goes unresubmitted.

$19.7B
Unnecessary denial-rework spend, US hospitals
Premier, 2022
11.8%
Initial denial rate, 2024 (10.2% in 2022)
Experian State of Claims
65%
Denied claims never resubmitted
MGMA Benchmarking

02 · The thesis

Adjudication is a structured-data problem masquerading as an AI problem

Every denial encodes a small disagreement between a clinical fact and a payer rule. The cost of resolving that disagreement is dominated by humans translating unstructured notes into the structured form payer logic actually needs.

Solve the translation deterministically and the rest of the workflow collapses. The agent does not need to be smart about most denials. It needs to be precise about the schema.

03 · The product

What it does

01

Schema-grade extraction

Convert encounter notes into a payer-aware structured representation. Every field carries field-level provenance back to the source.

02

Policy-aware reasoning

Match the structured claim against current payer policies and CPT/HCPCS guidance, with rule-level citations.

03

Auto-resubmit and appeal

Generate the resubmission package and electronic appeal payload directly from the resolution path.

04

Exception escalation

Route only the genuinely ambiguous denials to humans, with the model's prior chain-of-reasoning visible inline.

04 · Why now

The timing case

  1. 1

    Payer policy density has crossed the threshold where human-only adjudication cannot keep pace, and the gap widens every year.

  2. 2

    Foundation models are now reliable enough at structured extraction to anchor a deterministic downstream workflow.

  3. 3

    Provider revenue cycle teams are budget-frozen but losing more revenue per FTE every quarter. That is the forcing function for replacement, not augmentation.

05 · Why I see it

The view from inside the work

I led adjudication and patient-payment products at Emdeon and Change Healthcare, where denial economics drove every product decision on the platform. Workflow software does not solve this. The schema layer does.

06 · Comparable references

What's already in the market, and where the gap is

An honest read on the adjacent landscape. Not every comparable is a competitor. Some are partners. Some are the market the venture displaces.

Reference
Waystar / Availity
What they do
Workflow software for revenue cycle teams to manually work denials.
Gap
Sells productivity to humans. Apex sells the resolution itself.
Reference
Olive AI (defunct)
What they do
Tried full-stack RPA across health-system back office.
Gap
Bottomed out on a workflow-automation thesis. Apex is data-layer, not RPA.
Reference
Notable
What they do
AI scribes and intake automation in clinical settings.
Gap
Pre-claim, not post-denial. Apex starts where Notable ends.
Reference
In-house RCM teams
What they do
Outsource to manila and chase appeals manually.
Gap
Cost scales with denials. Apex breaks that linkage.

07 · Key risks

What could break the thesis

Operator-grade pre-mortem. Surfaced because the buyers and partners worth talking to will surface them anyway.

Risk · 01

Payer policies change weekly. The model becomes stale.

Mitigation

Schema is the product, not the model. Policy ingestion is automated and versioned, with rule-level provenance.

Risk · 02

Hospitals are budget-frozen and slow to buy.

Mitigation

Sell on contingency, structured to play clean with revenue recognition. The hospital CFO's first objection is auditor-shaped, not budget-shaped, and the contract has to be designed for that.

Risk · 03

Liability for an incorrectly resolved denial.

Mitigation

Auditable provenance for every decision. Customer keeps final-approval rights on edge cases. Insurance product on top.

08 · Proof of motion

What I've already shipped on this thesis

The artifacts that turn this from an essay into something with traction. Published work, working-group seats, operator scars.

09 · Questions partners ask

The next three follow-ups

Pre-empted because the buyers and partners worth talking to will surface them anyway.

Why doesn't a foundation model just solve this?

Because adjudication is a structured-data problem masquerading as an AI problem. Every denial encodes a small disagreement between a clinical fact and a payer rule. The cost of resolving it is dominated by humans translating unstructured notes into the structured form that payer logic actually needs. The agent doesn't need to be clever. It needs to be precise about the schema. Make the schema the product and the model is just an ingestion engine on top.

Hospital CFOs are budget-frozen. How do you actually sell this?

Contingency pricing, structured to play clean with revenue recognition. The hospital CFO's first objection is auditor-shaped, not budget-shaped. The contract has to be designed for that. Most contingency RCM contracts get torpedoed in revrec review, not in the budget meeting. Solve the auditor objection first.

Doesn't Olive AI prove this category is broken?

Olive AI tried full-stack RPA across health-system back office, with the workflow as the product. They sold ARR, scaled headcount, and bottomed out on the workflow-automation thesis. Apex is data-layer, not RPA. The product is the schema, not the dashboard. Different category bet entirely.

What's the first deployment look like?

A regional health system or RCM operation already running 11%+ initial denial rates, where the contingency math works for both sides on day one. We canonicalize their inbound claim data, run deterministic rule matching against current payer policies, and split the recovered revenue with rule-level provenance attached.

Status

Apex Adjudication is a published essay, not a stealth company. I am running Finexio. The thesis is here so the right operator or investor can find it and we can talk.

Of the eight ventures I've published, two are in discovery and I expect to operate one of them after Finexio. The rest, including this one, are pattern recognition I want in the open. If you read this and want to start it yourself, that is the outcome I'm hoping for.

Other concepts