skills.rest
nrakow
nrakowCOMMUNITY
Explore all skills by this author →
GitHub

entity-resolution

Resolve duplicate records across data sources into unified entity identities.

1|1|Updated Feb 27, 2026
Search Tags:#dbt#deduplication#data-quality#matching#entity-resolution#identity-graph#golden-record
One-click install
npx skills add https://github.com/nrakow/ae-skills-dev --skill entity-resolution-nrakow
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: entity-resolution
Source: https://github.com/nrakow/ae-skills-dev/tree/main/skills/entity-resolution
Command: npx skills add https://github.com/nrakow/ae-skills-dev --skill entity-resolution-nrakow

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Deduplicate and merge fragmented records across multiple source systems to produce a single, stable entity identity (golden record) for customers and entities.

Core Features & Use Cases

  • Deterministic exact-match clustering on emails across systems to quickly establish identity links.
  • Identity graph approach to propagate connections through transitive relationships (email, phone, IDs) for complex linked entities.
  • Optional fuzzy/name matching when identifiers are limited, with blocking keys to maintain performance.
  • Flexible golden-record construction to produce canonical attributes and an auditable trail of source_ids.
  • Confidence scoring and human-review workflows to balance automation with governance.

Quick Start

Validate your identity attributes in the data-stack-context and run the deterministic email-matching pattern to generate initial clusters.

Frequently Asked Questions about entity-resolution

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I deduplicate customer records across multiple data sources?▼

To deduplicate customer records across multiple sources, use deterministic exact-match clustering on emails and an identity graph to link transitive relationships, producing a stable unified entity identity.

What is the best way to build an identity graph for resolving duplicate entities?▼

Building an identity graph for entity resolution involves propagating connections through transitive relationships like email, phone, and system IDs to group fragmented records into a single cluster for canonical golden record construction.

Can I use dbt patterns for matching and merging fragmented customer data?▼

Yes, this approach provides dbt-ready patterns to resolve duplicate records across CRM, product, support, and billing systems, generating stable entity IDs and an auditable trail of source IDs for your data pipeline.

Does entity resolution work when source tables only have names and no common identifiers?▼

When source tables lack common identifiers, entity resolution can use optional fuzzy name matching combined with blocking keys to maintain performance while still linking records across your source systems.

How do you handle confidence scoring and human review for resolved identities?▼

Confidence scoring and human-review workflows balance automation with governance by validating matched entity clusters, ensuring data quality before propagating canonical attributes into your final golden record.

Related Skills

Explore More Data & Analytics Skills→
danbridanbri
1

skill-bill

Map UK Bill identifiers across historical and modern legislative data sources.

Community
Advanced
marquesfelipmarquesfelip

customer-offboarding

Automate end-to-end customer offboarding workflows for SaaS platforms.

Community
Advanced
armanzeroeightarmanzeroeight
29

test-generator

Generate schema, data quality, and freshness tests for dbt models.

Community
Intermediate
sunchenddsunchendd
3

wings-engine-patch

Patch AI inference engines at runtime via wrapt import hooks.

Community
Advanced