Cross-source identifiers
Every current identifier attachment, with the link method and confidence carried verbatim — deterministic links are 1.000; probabilistic links surface their real matcher confidence, never rounded up.
| Scheme | Value | Source table | Link method | Confidence | Since | Provenance |
|---|---|---|---|---|---|---|
| cms_op_profile_id | 3456990 | cms_open_payments | exact_npi_in_open_payments | 1.000 | 2026-06-13 | evidencerecorded: 2026-06-13 methodology: entity-graph/v1 evidence: {"npi":"1003234832","rule":"D-OP-NPI","recipient_profile_id":"3456990"} |
| npi | 1003234832 | nppes_providers | deterministic_pk | 1.000 |
Graph edges
One-hop relationships — ownership, chain membership, operation — each backed by a federal source row.
| Relationship | Entity | Confidence | Source table | Provenance |
|---|---|---|---|---|
| ← pays | Novartis Pharmaceuticals Corporation (NJ) | 1.000 | cms_open_payments | evidence |
Current record
Every current claim asserted about this entity, grouped by provenance subject. Each cell carries its upstream federal source.
provider · 1003234832 JSON
Point-in-time
Reconstruct this record as of any date, or diff it between two dates — both axes of the bitemporal store (bitemporal-read/v1): valid time (when a fact was true per the source) and recorded time (what the store knew when).
Pick a date to reconstruct, or two dates to diff. Equivalent JSON: /api/v1/history/{subject_type}/{subject_id}?at=… and …/diff?from=…&to=…