Medallion Architecture v1
Medallion Architecture v1
Medallion Architecture v1
Jun 2024
© 2024 Cognizant
Introduction
• A medallion architecture is a data design pattern used to logically organize data in a
Lakehouse, with the goal of incrementally and progressively improving the structure and
quality of data as it flows through each layer of the architecture.
5971308
Logical Architecture (Overall View)
Lake Ingestion Tier Lake Model & Transform Tier Serving Tier
Databases
Real-time Analytics
Global/Sector-Based
Metadata Driven Global/Sector-Based
Synapse SQL
Ingestion Environment SparkSQL Environment
APIs
Analytical
Applications
Centralized Presto
Store Store (today)
(future – decentralized)
Machine Learning
Disadvantages
• Uses large amounts of storage
• Often requires additional downstream processing
• Implies a data Lakehouse architecture
Store
Developers/
Leverage Azure to build, integrate, and manage data and analytics products. Create AI enabled applications / solutions when applicable.
Data Engineers
Data Analysts Leverage Azure to discover and share new insights from existing data assets or ad hoc data.
Data Scientists Use your preferred tools and machine learning frameworks to build scalable data science solutions. Accelerate end-to-end ML lifecycle.
Data Stewards Responsible for managing access to data assets. Authorization control – approve access to data sets or delegate that approval accordingly.
6
Thank you