Our data architecture follows a structured, layered approach based on the Medallion model. It aligns with Microsoft Fabric principles and best practices, supporting efficient data ingestion, transformation, governance, and reporting.
Landing Layer
This layer receives raw data files (typically CSV format) from external systems into secure landing zones.
Bronze Layer
Bronze provides the initial structured storage of raw data into Parquet format within the Lakehouse, forming the immutable ingestion layer.
Silver Layer
Silver refines and transforms the data, applying business rules, historical comparisons, and creating utility views ready for wide consumption.
Gold Layer
Gold provides business-aligned semantic models, including CDM Commodity and Utility Views, supporting reporting, analytics, and Fabric datasets.
Release and Archive Layers
We manage data lifecycle with snapshot queries, versioned releases, and archive pathways to ensure compliance, auditability, and scalability.
Attribute Lineage (Data Dictionary)
A new feature in the model is the inclusion of an Attribute Lineage and Data Dictionary workflow.
This tracks metadata and field-level lineage across the data layers, providing enhanced traceability and documentation, feeding into the development pipelines (VSCode and DevOps).