One canonical identity. Every federal source.

NPPES uses NPIs. PECOS uses PECOS-IDs. CMS Care Compare uses CCNs. OIG LEIE uses NPIs and legacy LEIE-IDs. HRSA HPSA uses Provider-IDs. Fonteum resolves them into one canonical identity per provider across all 5 federal sources, with confidence-banded matching + per-link methodology version + source evidence sufficient to manually audit every link decision.

0providers canonically linked across 5 federal sources·0high-confidence links·methodology —·last extended —

Verifier reference + code samples → Cryptographic chain → Snapshot integrity →

Confidence distribution

Public endpoints default to high (≥ 0.9) only.

No identity links yet. The linker runs every 5 minutes against the NPPES specialty caches + 4 source-side tables (pecos_providers, carecompare_providers, leie_providers, hrsa_providers). All 4 source tables ship empty; once Cowork backfills normalize source_snapshots payloads into them, the linker will begin producing links and this chart will populate.

Match methods

Nine strategies across 5 federal sources, confidence-ordered.

NPPES ↔ PECOS

exact_npi_in_pecos · confidence 1.0
PECOS row carries an explicit npi matching an NPPES record. Strongest possible link.
exact_name_address · confidence 0.95
Normalized provider name AND byte-equal normalized address. Not 1.0 because group practices can share an office.
name_address_fuzzy · confidence 0.7-0.9
Jaro-Winkler similarity, gated by exact state + zip5. Fuzzy ceiling 0.9 — public default (≥ 0.9) excludes bottom of fuzzy band.

NPPES ↔ Care Compare CCN

exact_npi_in_carecompare · confidence 1.0
Care Compare row carries an NPI matching the NPPES record.
exact_name_address_carecompare · confidence 0.95
Facility name + byte-equal address. No fuzzy strategy — Care Compare facility names have high lexical variance; deferred to Phase 3.

NPPES ↔ OIG LEIE

Sample queries

Three reads. No credentials.

# Resolve an NPI to its canonical PECOS-ID
curl https://fonteum.com/api/v1/identity/nppes/1245319599

# Resolve a PECOS-ID to its canonical NPI
curl https://fonteum.com/api/v1/identity/pecos/PAC-100001

# Aggregate stats: total linked + confidence-band distribution
curl https://fonteum.com/api/v1/identity/stats

# Opt-in to surface low-confidence (<0.9) links for diligence
curl 'https://fonteum.com/api/v1/identity/nppes/1245319599?include_low_confidence=true'

Full reference + Node + Python + Go samples at /docs/identity.

The Bloomberg ticker for healthcare providers

Once a canonical identity layer exists and is queryable, it becomes the default reference for the category. Every Fonteum consumer (MCP server, webhook events, semantic search, future dbt package) joins against provider_identity_links, not against the federal sources directly.

Zenodo

DOI 10.5281/zenodo.14872055
deposited 2026-05-08

medRxiv

preprint 2026.05.08.26337284
methodology paper · v1.0

Hugging Face

dataset · 2.1 GB
6.8M+ records · NDJSON

DuckDB

public drop · parquet
read-only · v2026.05.0

GitHub

provider-graph-spec
v1.0 · MIT licensed

Compliance posture

Compliance posture

One canonical identity. Every federal source.

Public endpoints default to high (≥ 0.9) only.

Nine strategies across 5 federal sources, confidence-ordered.

Three reads. No credentials.