Reltio Embedded Entity Resolution in Databricks at a glance
Learn about Reltio Embedded Entity Resolution in Databricks, a Databricks-native solution that identifies and groups matching individual records directly in your Databricks environment.
Reltio Embedded Entity Resolution for Databricks is a Databricks-native solution that identifies and groups matching Individual entity records directly within your Databricks environment, without requiring any data movement or duplication. It operates entirely on data stored in your Unity Catalog, ensuring processing stays within your Databricks account.
Reltio Embedded Entity Resolution uses Reltio's Flexible Entity Resolution Network (FERN) models to identify matching records and group them together at scale. The solution generates match scores between the grouped records. These match groups can support downstream analytics, machine learning workloads, and operational processes that depend on consistent individual data.
Who can use this solution
Reltio Embedded Entity Resolution is designed for organizations that manage large volumes of individual data in Databricks and need to identify duplicate or matching records. It's intended for the following user roles:
Data Product Owner
Solution Architect
Data Steward
Developer
System Administrator
Benefits of Reltio Embedded Entity Resolution
Individual data often remains fragmented across catalogs, schemas, and tables, even when it is stored in Databricks. Records that represent the same person can appear multiple times with differences in names, contact details, addresses, or identifiers.
With Reltio Embedded Entity Resolution, you can:
- Identify duplicate or related individual records.
- Create consistent entity groupings across datasets.
- Improve the accuracy of analytics and machine learning models.
- Support reliable reporting and decision-making.
Reltio Embedded Entity Resolution helps you create more consistent entity groupings while keeping resolution processing in your Databricks environment.
Key capabilities
- AI-powered matching in Databricks: Uses pretrained Flexible Entity Resolution Network (FERN) models to identify and group records that refer to the same individual entity across one or more datasets.
- Native execution in Databricks: Runs entirely within your Databricks environment and works with Delta Lake, Unity Catalog, and Databricks compute resources.
- Configurable matching: Allows you to map source columns to standard individual entity attributes and adjust matching behavior using predefined or custom configuration settings.
- Match groups and scores: Generates grouped records with a stable group ID and match score, enabling analysis, validation, and downstream processing.
- Interactive result exploration: Enables you to review matched records, inspect group details, and analyze match outcomes directly in Databricks.
Where to access the solution
You can access Reltio Embedded Entity Resolution in Databricks on the Databricks Marketplace.
For setup instructions, see Onboarding setup workflow.