Data & Analytics Architect — Microsoft Fabric & Power BI
Experience:
10+ years (3+ in Fabric or equivalent Azure data platform) Role Summary You are the end-to-end technical authority for our analytics estate — from Oracle/JD Edwards source systems through ingestion, Lakehouse and semantic layer, to certified Power BI data products.
The role
spans the full Microsoft Fabric stack and enterprise Power BI architecture, with a hands-on build component alongside the design work.
What You'll Do
- Own the target-state Fabric architecture: workspace/domain topology, OneLake structure, medallion layering, and Lakehouse vs.
- Warehouse vs.
- Eventhouse decisions.
- Design and build Fabric Data Factory pipelines and Spark notebooks (PySpark/Spark SQL) metadata-driven ingestion, orchestration, Delta table optimisation, error handling and monitoring.
- Architect ingestion from Oracle / JD Edwards: incremental and CDC extraction, gateway connectivity, reconciliation back to ERP source-of-truth.
- Architect enterprise semantic models — star schemas, Direct Lake vs.
- Import vs.
- Direct Query, advanced DAX, performance tuning at scale.
- Define Power BI governance: RLS/OLS, Entra ID group strategy, certified datasets, workspace and tenant settings, gateways, licensing.
- Establish CI/CD via Fabric Git integration and deployment pipelines, own Fabric capacity sizing, monitoring and cost optimization.
- Set standards, review code and architecture, and mentor engineers, BI developers and analysts.
- Must Have Microsoft Fabric, full stack, hands-on: One Lake, Lakehouse/Delta, Warehouse, Data Factory pipelines, Dataflows Gen2, Notebooks, semantic models, Direct Lake.
- Power BI at architect level (6+ yrs): dimensional modelling, advanced DAX and optimization, RLS/OLS, enterprise governance and ALM.
- Oracle Database: strong SQL and PL/SQL, query tuning, and proven extraction from large normalized ERP schemas.
- Advanced SQL, solid Python, Spark tuning, Git-based CI/CD.
- Azure fundamentals (Entra ID, ADLS Gen2, Key Vault) and Fabric admin/security model.
- Clear communication with both engineering and business stakeholders; strong documentation discipline.
- Good to Have JD Edwards EnterpriseOne functional knowledge — Finance, Distribution and/or Manufacturing; familiarity with F-tables (F0911, F4211, F0411), UDCs, Julian dates and JDE data conventions.
- Microsoft certifications : DP-600 / DP-700, PL-300.
- Experience migrating legacy data and analytics platforms to Fabric
