Golden Lake by Entrada

Golden Lake is a Databricks-native master data management accelerator built by Entrada that delivers probabilistic entity resolution, inline data quality remediation, and human-in-the-loop stewardship. It produces compliance-ready golden records directly on Delta tables with no external MDM platform.

The accelerator unifies fragmented customer and vendor records from CRM, ERP, billing, and 130+ SaaS systems using EM-trained comparison weights instead of brittle deterministic rules. It auto-nullifies placeholder values, malformed emails, and defaulted dates before survivorship, then routes ambiguous matches to data stewards through a Streamlit-powered Databricks App, feeding every decision back into the matching model for continuous learning.

Benefits of the Golden Lake Accelerator include:

• Resolves duplicates across CRM, ERP, and billing using EM-trained matching

• Runs entirely on Delta tables and Unity Catalog (no external MDM platform)

• Routes ambiguous matches to stewards with full audit trail and model feedback