The Data Backbone

One rate-intelligence estate.Every payer, every code,every market.

~55 federal, commercial, and geographic datasets, fused into one normalized graph — joined on NPI, CPT, and geography — that scores any provider's rate against its true local peers and turns every rate into one comparable number: % of Medicare.

Explore the estate ↓ See what it powers →
0
federal rate records
0
payers
0
states
0
provider NPIs
0
Medicare codes normalized
0
datasets fused

Scope of the public estate we continuously index toward — not a live row count. Coverage by code, payer, and market is disclosed on every result. See our honesty standard →

The one number
The problem

Every buyer and every competitor speaks in "% of Medicare" — but nobody had a clean Medicare denominator for the down-market layer.

What we built

A locality-correct Medicare engine.

Work / PE / MP RVUs × locality GPCIs × the conversion factor → a locality-correct Medicare allowed amount for every code, served by the medicare_allowed() engine.

19,356 codes → locality-correct Medicare allowed → % of Medicare on any rate.
verified · 99213 national $95.19 · 110 GPCI localities
Why no one else does this

The moat is integration,
not raw size.

Nine data domains in one normalized graph — commercial TiC, out-of-network, Medicare benchmarks, drug pricing, hospital charges, the provider/group graph, geography, quality, and discovery — joined on NPI, CPT, and market. Every other vendor brings two to four, siloed by delivery format. We normalize ~55 datasets to the same 110 GPCI localities on a $0 public-data cost base, so a single CPT speaks as a % of Medicare, a commercial rate, a hospital charge, and a Medicaid fee at once.

Local-peer scoring is the payoff of that fusion — and it is being hardened from national to true metro-level accuracy. We lead with nine domains in one graph (true today) and frame local-peer scoring as the engine that breadth makes possible, not a production-proven metro number.

2–4× the domain count of any rival
9 domains
Trilliant, the broadest rival, joins four — and ships them siloed by format. Payerset, Clarify, and Mathematica land around three; Serif, SimpleHC, and PayerBenchmark stop at two. Reddenda joins nine on shared keys: 314M+ federal records compressed into one queryable graph across 19,356 normalized Medicare codes and 935 metros.
Integration, not raw files
110 localities
Rivals gate these sources behind $75K–$350K contracts and deliver them as separate dashboards, APIs, and feeds — or through consultants who do not scale down-market. We normalize all nine to the same 110 GPCI localities, then finish the work: a send-ready Leverage Memo at $199, a category they structurally cannot reach because they stop at data.
Local-peer is the payoff
$0 data cost
Scoring a rate per-NPI, per-CPT against its true local peers needs the Medicare RVU engine, commercial TiC, the provider graph, and geography working together. That is exactly why no incumbent markets it down-market — and why the fusion, not the raw size, is the asset.
The estate

Browse the data backbone.

One normalized graph. Filter by what you care about — domain, what's live, the tool it powers, the audience it serves.

Quarterly fee-schedule pulls · monthly registry & payer-index diffs · rolling rate-stat rebuilds. Wired to refresh, not frozen.
19,356 Medicare codes 935 metros mapped 82,817 groups in the affiliation graph 701 canonical specialties 9.2M+ provider NPIs
Live in production today Indexed landed, wiring in flight Expanding acquiring & deepening
Showing 38 of 38 dataset families · ~55 underlying sources (siblings bundled)
Data → tools

What the estate powers.

Every dataset above feeds a live surface. The data doesn't sit in a warehouse — it scores rates, builds memos, and maps markets.

Web
RateScore
Your rate vs your true local peers, as one 300–850 score.
Commercial TiC · RateScore engine · PFS · NPPES · CBSA · specialty
Powered by 6 datasets →
Try it free →
Memo
Leverage Memo
A documented, defensible negotiation target per code.
Commercial TiC · OON · PFS · ASP · reference benchmarks
Powered by 5 datasets →
Order a Leverage Memo →
Web
Whitespace
Where the upside is — panel size × acuity = yield.
Part B NPI-profile · Geographic Variation · ACS · HPSA · PLACES
Powered by 5 datasets →
Find your upside →
Web
DME Rate Intel
Per-NPI local-peer DME scoring + "% of DMEPOS".
DMEPOS · Commercial TiC (DME codes) · NPPES
Powered by 3 datasets →
See DME scoring →
In-app
Atlas (app)
Multi-NPI / group rollups for MSOs and billers.
DAC group graph · reassignment edges · Part B · TiC
Powered by 4 datasets →
Open Atlas →
Web
Free NPI Lookup
Look up any provider's profile + benchmark, free.
NPPES · procedure volume · Medicare Part B
Powered by 3 datasets →
Look up an NPI →
One backbone, three lenses

Built for the people who use the data.

0provider NPIs in the market
0Medicare codes normalized
0datasets fused
Know exactly what every payer pays for your codes — scored against the providers around you, with a documented target. Built on public data. No PHI.

You already suspect you're underpaid on some codes and fine on others — you just can't prove which, or by how much. We turn the public rate estate into your answer: for each of your codes, your rate next to your local same-specialty peer set, expressed as a percent of Medicare, with the gap quantified and the target documented.

Local-peer, not national
Per-NPI, per-CPT, per-metro, per-specialty — a urologist compared to urologists, in their own market.
One number
% of Medicare on every rate, across 19,356 normalized codes.
No PHI
Public data only. Nothing to upload, nothing to expose.
0federal rate records
$0data cost
0domains in one graph
Enterprise-grade rate intelligence — the kind that sells for $75K–$350K per engagement — built on $0 of data cost, pointed at the segment incumbents can't serve profitably.

The moat is a compounding public-data estate: ~55 federal, commercial, and geographic datasets fused into one normalized graph, refreshed on a cadence, with a per-NPI local-peer engine no incumbent offers down-market. The architecture keeps it cheap — a built data-lake offloads the heavy raw; compact rollups serve the live site.

Parity at zero data cost
The same public sources the enterprise vendors gate behind six-figure contracts.
Defensibility
The normalization + local-peer engine — not the raw files — is the asset.
Operating leverage
Hetzner data-lake + compact Supabase rollups; the heavy lifting never touches the serving tier.
0groups in the graph
0clinician affiliations
0metros mapped
Build on the estate — group and MSO rollups, white-label benchmarks, and API-grade rate intelligence for your whole book of providers.

Bring your providers; we bring the backbone. The group graph — 82,817 groups and 1.76M clinician affiliations — rolls a whole MSO or billing book into one view, every NPI scored against its real local peers, every memo white-labelable under your brand.

Group / MSO rollups
Multi-NPI Atlas across your entire book in one view.
White-label
Partner Leverage Memos under your brand.
Co-built coverage
Your specialties and markets prioritized in the expansion.
Provenance is the proof

Honest by construction.

Coverage transparency — every result discloses availability by code, payer, and market.
Public & licensed sources only
Federal MRFs, CMS catalogs, the NPI Registry, public geographic data. No scraped PHI; nothing you upload.
Coverage disclosed on every result
We tell you, per result, how deep the data is by code, payer, and market. Our standard →
Wired to refresh
Quarterly fee-schedule pulls, monthly registry & payer-index diffs, rolling rate-stat rebuilds.
One normalized graph
Joined on NPI, CPT, and geography — a commercial rate, a Medicare allowed, a hospital charge, and a Medicaid fee all speak the same language.

One backbone. Score your rates against the real market.

See your rate against your true local peers — built on the public estate above, in about a minute.

Run a Free Snapshot →

Generated in about 15 seconds. No email. No credit card.

Every number here is the scope of the public estate we index toward — not a live row count. Coverage is disclosed on every result. Methodology & sourcing → · Our honesty standard →