The Evidence Engine

4.8 million payment disputes.
The evidence that wins them.

Since 2022, federal law lets a provider take an underpaid out-of-network claim to binding arbitration. Roughly 2.6 million disputes were filed in 2025 alone, and providers won 88% of the decisions. The law tells the arbitrator exactly what to weigh: the provider's own contracted-rate history and local market share. That is our database.

45 CFR §149.510(c)(4)(iii)(B): the certified arbitrator must consider the contracted rates between that provider and that plan during the previous 4 plan years, and the market share held by the provider or the plan in the geographic region. Reddenda's index is built from the public record of exactly those two things.

The market, in CMS's own numbers

This is not a projection. It is the federal docket.

4.8M+federal IDR disputes filed since the process opened in April 2022CMS public use files, via Georgetown CHIR
~2.6Mdisputes filed in calendar 2025 alone; H1 2025 was up 39% over H2 2024CMS Supplemental Background PUF
88%of federal IDR determinations in H1 2025 were won by the provider sideCMS data; up from 81% in 2023
$2.2–2.8Bflowing to providers per year in additional payments through the processmodeled from CMS volumes; $2.24B measured across 2023–24 determinations

Honest footnote: the 88% is the aggregate CMS rate, and it is driven by sophisticated repeat filers — the ten largest initiating parties file roughly 70% of all disputes and win at 82–95%. That is precisely the point. Winning is an evidence discipline, the biggest filers industrialized it, and the long tail of independent practices has had no instrument. Nothing on this page is a guarantee of any outcome.

Why Reddenda wins this category

The statute names the evidence. We hold it.

Every dispute turns on the same question: what should this claim have paid? The payer answers with the QPA, its own median of its own rates. The arbitrator is instructed not to double-count it — and is required to consider what the provider submits about the factors the QPA cannot contain.

Factor (B)(5)

Contracted-rate history

The rates between that provider and that plan over the previous four plan years. Reddenda's index is built from the public Transparency-in-Coverage record of contracted rates — per NPI, per code, per payer — and every new filing compounds it.

Factor (B)(2)

Local market share

The market share held by the provider or the plan in the geographic region. Our engine already computes local peer distributions for every score it issues; the same machinery renders the market-share exhibit for the region in dispute.

The one-sentence version: federal law made the winning evidence a matter of public record, and Reddenda is the engine that reads that record. The exhibit a $75,000-a-year enterprise platform would assemble for a hospital, we produce for an independent practice — certified, tamper-evident, and priced like a filing fee.

The gap we fill

The provider's instruments changed hands. Nobody armed the long tail.

12 JAN 2026
Zelis acquired Rivet.
The leading independent provider-side underpayment-analytics tool is now owned by a healthcare-payments company that serves more than 750 payers, including the five largest national plans. Whatever Rivet becomes next, its independence is over. Zelis newsroom · BusinessWire, 12 Jan 2026
24 JUN 2026
Zelis launched AI arbitration tooling for payers.
An AI-native product built to help payers assemble stronger No Surprises Act briefs and improve their win rate. A saving to the plan is a dollar the provider was not paid. Zelis newsroom · BusinessWire, 24 Jun 2026
TODAY
Litigation analytics serve law firms. The QPA serves payers.
Trellis-class court analytics are priced and built for legal teams, not practices. The dispute process itself hands the payer a default number. The independent practice walks into a federal proceeding with the least tooling of anyone in the room.

Our covenant: Reddenda has never taken a dollar from a payer, and never will. Every exhibit we build is built for one side of the table. Read the covenant →

How it works

From an NPI to a filed exhibit.

1

The index already knows you

Run any NPI. The engine holds the public contracted-rate record and computes the local peer distribution for the region and codes in dispute. Looking is free, and no PHI is ever required.

Run a free rate card →
2

The exhibit, built to the statute

A certified, tamper-evident evidence exhibit prepared for IDR submission: your current contracted-rate position and the local market-share exhibit, source-cited to the public federal filings, structured to the factors the arbitrator must weigh.

IDR Evidence Exhibit · $149 per dispute
3

At scale: the performance lane

For qualifying practices and billing companies we build the evidence book across the whole dispute pipeline and price it as a share of documented recovery — only where state law permits a performance fee, and always flat-fee in states where it does not. Scoped on a call, closed with a signed engagement letter.

Share of documented recovery · by state law
The revenue engine, sized honestly

A percentage of a measured river.

Every input below is a published CMS figure or is derived from one, and the derivation is shown. The addressable segment is the long tail the industrialized filers ignore: the ~30% of dispute volume filed outside the ten largest initiating parties.

LayerBasisAt 10% takeAt 15% takeAt 25% take
Whole market (ceiling) ~2.6M disputes/yr · 88% win · $2.2–2.8B/yr in awards $220–280M/yr$330–420M/yr$550–700M/yr
Addressable long tail ~780K disputes/yr outside the top-10 filers · $650–850M/yr in awards $65–85M/yr$100–130M/yr$165–215M/yr

Modeled from CMS-published dispute volumes and measured award flows (Georgetown CHIR analysis of CMS public use files). These are market-size layers, not revenue projections or guarantees; capture depends on execution, state-law eligibility for performance fees, and win outcomes that are never guaranteed. The exhibit itself is also sold flat-fee at $149 per dispute, so the engine earns in both lanes.

Where it sits in the revenue architecture

$0 RateScore $299 Memo $697 Audit $399/mo Pro $149/dispute Exhibit % of recovery, at scale the data platform →

One free viral wedge feeds a flat-fee ladder; the ladder feeds the evidence engine; every exhibit and every score compounds the same index. The platform layer — the mass rate-data infrastructure — is the moat under all of it.

The honest spine: every dollar figure on this page is a documented, modeled figure from public federal data, never a guaranteed recovery. The 88% is the aggregate CMS determination rate for H1 2025, not a promise about any dispute. Performance fees are offered only where state fee-splitting law permits them; New York, Illinois, and North Carolina are always served flat-fee. Eligibility is verified server-side against the NPPES practice record. No PHI is required for any of it.

The other side of the table is already armed

Bring the record.