Medallion Architecture Layer - Silver

Medallion Architecture Layer - Silver


The Silver Layer is the intermediate stage in the Medallion Architecture where raw data is filtered, matched, merged, and cleansed, applying "just-enough" transformations to refine the dataset. This layer establishes the "Enterprise view" of key business entities and transactions, serving as the first stable layer in the data pipeline. Its functions include cleaning corrupt rows, implementing data quality constraints, and augmenting data by joining across other tables, setting the foundation for advanced analytics and modeling.



Data Trust and Consistency: The Silver layer ensures a higher degree of consistency and reliability by performing cleaning and filtering on the raw Bronze data, preventing unexpected or corrupt rows (like the _corrupt column) from progressing downstream.

Foundation for Advanced Analytics: This layer provides the normalized, cleansed input data necessary for use cases such as self-service analytics, ad-hoc reporting, advanced analytics, and training data for AI/ML models.

Enables Data Augmentation: The Silver layer can be used to augment data by joining data from one or many tables—including other Silver tables or even Gold tables—to produce new refined datasets, facilitating the creation of key business entities.





In the Silver layer of the lakehouse, the data from the Bronze layer is matched, merged, conformed and cleansed ("just-enough") so that the Silver layer can provide an "Enterprise view" of all its key business entities, concepts and transactions. (e.g. master customers, stores, non-duplicated transactions and cross-reference tables).



Source: https://delta.io/pdfs/dldg_databricks.pdfPage: 205 • Key: leeDeltaLakeDefinitive