Healthcare Intelligence Platform
End-to-end data engineering and analytics platform for a multi-facility senior care operator — from enterprise source systems to real-time executive dashboards.
A multi-facility senior care organization was operating across disconnected source systems for clinical data, workforce management, HR, and vendor payments. Leadership had no unified view of census trends, revenue projections, AR aging, or workforce costs. Reports were manual, delayed, and inconsistent across facilities.
Designed and deployed a full Microsoft Fabric data platform — ingesting from 4 enterprise source systems through secure database connectivity, processing through PySpark notebooks and Dataflow Gen2 transformations, and delivering a 144-measure Power BI semantic model with real-time visibility across all facilities.
What Was Built
A production-grade Fabric platform with full pipeline orchestration, secure ERP ingestion, PySpark transformations, and a 144-measure semantic model serving C-suite and facility leadership.
Secure Enterprise Integration
Wallet-based JDBC connection with .p12, .pem, and tnsnames.ora credentials — enabling secure ingestion from a restricted healthcare ERP into Microsoft Fabric.
Lakehouse & Delta Layer
A centralised lakehouse as the single source of truth — structured Delta tables powering all downstream analytics with full auditability.
10 Automated Pipelines
End-to-end orchestration for census, revenue, labor, RCM, vendor AP, AR aging rollup, and semantic model refresh — fully scheduled and monitored.
18 PySpark Notebooks
Data cleansing, resident joins, HR processing, open position tracking, anticipated discharge payer classification, and Gemini AI feature engineering.
Dataflow Gen2
Power Query transformations for census data, revenue cycle management, and row-level security user mapping.
144-Measure Semantic Model
DAX model covering census, revenue projection, AR aging, workforce PPD, leads conversion, Medicaid pending, and DSO tracking.
Results & Outcomes
- Real-time census, revenue, and workforce visibility delivered to C-suite and facility leadership for the first time
- AR aging tracked across 13 buckets (Current through 360+ days) with dynamic date anchoring
- Projected month-end revenue modeled day-by-day using carry-forward census logic, differentiated by care type
- Workforce PPD benchmarked against budget with OT variance, open position, and light duty tracking
- Row-Level Security enforced across multi-facility user roles
- Gemini AI integrated via notebook for predictive feature engineering
Technologies Used
Need a similar healthcare data platform?
Let's discuss how we can unify your operational data and give leadership real-time visibility into what matters most.
Work with Me