Data Warehouse Cloud Migration

A healthcare company operated a legacy on-premise data warehouse based on SQL Server. The existing solution reached its limits: growing data volumes, lack of self-service capabilities for departments and high maintenance effort. The requirement was a complete cloud migration to a modern, scalable data platform.

Phase 1 – Architecture

The project spanned two phases: In the first phase, the target architecture was designed. This included evaluating various cloud data platforms, deciding on Microsoft Fabric as the target platform, and developing a comprehensive cloud and data architecture. Conventions, best practices and standardized, metadata-driven data pipelines were defined, enabling consistent data processing across all source systems.

Phase 2 – Migration

In the second phase, the migration itself took place. Building the Microsoft Fabric platform involved setting up workspaces, lakehouses and deployment pipelines. The data pipelines were developed metadata-driven – new data sources can be added through configuration instead of code. In parallel, DevOps pipelines for automated testing and deployment were established.

Master Data Management

Additionally, a central Master Data Management based on Power Apps was developed. This enables maintenance of master data and control of Row-Level Security (RLS) for permission management in Power BI. Another low-code application for project time tracking and reporting was created, including a bidirectional ERP interface for booking and invoicing project times.

Reporting

The reporting landscape includes various Power BI reports and ad-hoc analyses for finance, HR, IT, project consulting, project management and monitoring – each with granular permission concepts.