On-premise Data Warehouse migration to Microsoft Fabric

Problem

An on-premise Data Warehouse limited scalability and modernization. Modeling, pipelines, SQL procedures and reports had to move to the cloud without losing consistency.

Architecture

flowchart LR
  DW[(On-Premise DW)] --> MIG[Migration]
  MIG --> F[Microsoft Fabric / OneLake]
  F --> LH[Lakehouse]
  LH --> PROC[Pipelines + PySpark]
  PROC --> SQLW[Synapse DW / T-SQL]
  SQLW --> PBI[Power BI]
  classDef old fill:#1a1018,stroke:#f59e0b,color:#e2e8f0
  classDef new fill:#0d1525,stroke:#10b981,color:#e2e8f0
  class DW,MIG old
  class F,LH,PROC,SQLW,PBI new

Solution

• Dimensional modeling and schema optimization on the new platform.
• Built and tuned ingestion and transformation pipelines with Fabric.
• Migrated and re-implemented SQL procedures to the Lakehouse environment.
• Published and tuned Power BI reports and dashboards connected to the Lakehouse.

Results

• Modern, scalable, cloud-based platform.
• Standardized ingestion and transformation pipelines on Fabric.
• Reports served directly from the Lakehouse.

Stack