Architect Guide

Welcome

Welcome to the Golden Core Architecture Guide. This documentation suite is designed for external users, system integrators, and data stewards who need to understand and work with the Golden platform.

Golden Core is a master data management and entity resolution platform that provides intelligent data deduplication, search capabilities, and golden record creation.

This guide is organized into six focused documents, each covering a key architectural component:

System Diagram

Architecture Diagram

Quick Start Paths

Recommended reading order:

• Entities Guide - Understand entity types and configuration

• Resources Guide - Learn about datasets, indexers, classifiers

• Golden Records Guide - Master deduplication and golden records

• Tables Guide - Understand data storage layer

Key Concepts:

• Data

  • Entity types (NONE, SEARCH, DUPLICATES, AUTO_DUPLICATES)

  • Dataset schema definition

  • Indexer configuration for duplicate detection

  • Classifier and merger strategies

• Security

  • API authentication with JWT tokens

  • Role-based access control (RBAC)

  • Built-in roles (ADMIN, STEWARD, USER)

  • Permission system and wildcards

  • Access token management

• Automations

  • Bucket classification (MATCH, REVIEW, NON_MATCH)

  • Manual vs automated duplicate resolution

  • Record quality scoring

  • Bucket operations (merge, split, disconnect)

Core Concepts Overview

Entities

Entities are the primary abstraction for managing master data:

• Wrap tables with intelligent features

• Support search via indexing

• Enable duplicate detection and resolution

• Orchestrate data sources and sinks

• Provide automatic or manual synchronization

Types: NONE → SEARCH → DUPLICATES → AUTO_DUPLICATES

Resources

Resources are reusable configuration components:

• Datasets - Define data structure/schema

• Sources - Read from external systems (files, databases, APIs)

• Sinks - Write to external systems (Kafka, HTTP, databases)

• Transformations - Map between data structures

• Pipelines - Multi-stage data processing

Golden Records

Golden Records represent the "single source of truth":

• Records - Individual data documents with metadata

• Buckets - Groups of potentially duplicate records

• Classification - Similarity scoring (MATCH, REVIEW, NON_MATCH)

• Merging - Creating master records from duplicates

• Quality Scoring - Data completeness and validity metrics

Tables

Tables provide persistent data storage:

• Store records in PostgreSQL

• Reference dataset for structure

• Support history tracking

• Provide filtering and pagination

• Enable ETL operations (load, transform, export)

Security

Multi-layered security model:

• Authentication - Username/password, JWT, SSO (Google, Azure)

• Authorization - Role-based access control (RBAC)

• Permissions - Fine-grained operation control

• Entitlements - Data-level filtering

• Access Tokens - Programmatic API access

Tasks

Asynchronous background job system:

• Entity Tasks - Synchronization, loading, indexing

• ETL Tasks - Load, transform, export operations

• Scheduling - One-off or CRON-based

• Monitoring - Real-time progress tracking

• Notifications - Email alerts on completion

API Documentation

Golden Core provides comprehensive REST API documentation:

Swagger UI: Access interactive API documentation at /swagger-ui

For more info, review the Developer Guide section

Getting Help

• Check relevant guide troubleshooting section

• Review Swagger API documentation

• Examine task logs for detailed errors

• Check system health endpoint: /actuator/health

• Contact system administrator for support