parent-map

What problem does it solve?

Consolidates disparate company identifiers and messy name variants into a single, code-driven parent company reference so downstream dives and gold models join consistently and market metrics remain accurate.

Core Features & Use Cases

• Code-first joins: Resolve entities by code + code_type (NAIC group/cocode, MA contract, CMS CCN, etc.) rather than fuzzy name matching to ensure deterministic joins.
• Enrichment + reconciliation: Batch unmapped codes through parallel.ai, collapse failures to self-named entities, canonicalize parent names, and tag network affiliations (BCBS, Delta Dental, Kaiser).
• Audit and publish: Produce a reconciled CSV, upload via soria scraper flow, enforce a 100% match verification gate, and wire the deduplicated mapping into gold models.
• Use Case: Turn a newly ingested NAIC dataset with mixed group and cocode entries into a single parent_company column in dives with historical name variant support.

Quick Start

Ask parent-map to identify unmapped codes from the target source, run parallel.ai enrichment, reconcile canonical parents and produce a reconciled CSV ready for soria scraper upload and warehouse publish.