Analytics Bi
Executive Healthcare BI System
From healthcare operations data to executive decision support.
A healthcare analytics and business intelligence system that prepares and validates source data in Python, models it dimensionally for analysis, and delivers executive, operational, clinical, demographic, and financial insights through SQL and Power BI.
Problem
Context
Healthcare operational data spans patients, admissions, departments, dates, clinical outcomes, and financial measures that must be modeled consistently before leaders can use it for decision support.
Challenge
Transform source healthcare data into a validated analytical model while preserving clear metric definitions and separating operational measures from assumptions or proxy metrics.
Objective
Build an executive BI system with reproducible data preparation, dimensional modeling, SQL analytics, validated KPIs, and Power BI reporting.
Architecture
Python prepares and validates healthcare source data, a dimensional model organizes analytical entities, SQL supports analytical querying, and Power BI provides the semantic and dashboard layers for executive reporting.
- 01
Healthcare source data
- 02
Python preparation and validation
- 03
Dimensional analytical model
- 04
SQL and semantic measures
- 05
Power BI executive dashboards
Data Preparation and Validation
Cleans, transforms, validates, and prepares healthcare source data for analytical modeling.
- Python
- Pandas
Dimensional Model
Organizes admissions, patients, departments, and dates into fact and dimension tables for consistent analysis.
- SQL
Analytics Layer
Supports analytical queries and KPI validation against prepared healthcare datasets.
- SQL
- Python
Semantic Model
Defines business measures and analytical relationships for dashboard reporting.
- Power BI
- DAX
Executive Dashboards
Presents executive, operational, clinical, demographic, and financial views of healthcare performance.
- Power BI
Technology
Language
- Python
- SQL
- DAX
Data
- Pandas
- Dimensional Modeling
Analytics
- Power BI
Visualization
- Power BI Dashboards
Engineering Contribution
Built a validated healthcare analytics pipeline
Prepared and validated healthcare source data in Python before analytical modeling and reporting.
Designed the dimensional analytical model
Structured healthcare analytics around admissions, patients, departments, and dates using fact and dimension tables.
Implemented executive BI reporting
Built Power BI reporting across executive, operational, clinical, demographic, and financial perspectives.
Validated key healthcare metrics
Verified analytical outputs and documented important metric limitations rather than presenting proxy measures as stronger claims.
Engineering Evidence
Code
Python Data Preparation and Validation
The repository contains Python workflows for preparing healthcare data, generating KPIs, building the date dimension, and producing validation outputs.
Architecture
Healthcare Dimensional Model
The analytical model includes fact_admissions with dim_patients, dim_departments, and dim_dates.
Validation
Validated Analytical Dataset
Validated outputs include 101,766 admissions, 71,518 patients, 73 departments, and 3,287 dates.
Metric
Verified Healthcare KPIs
Verified outputs include a 46.09% readmission rate, 4.40 average length of stay, and 3,474.27 average treatment-cost proxy.
Dashboard
Executive Power BI Reporting
The BI layer presents executive, operational, clinical, demographic, and financial views of healthcare performance.
Documentation
Documented Metric Limitations
Project documentation identifies treatment cost as a proxy metric and does not characterize the readmission measure as a specific 30-day readmission rate.
What It Proves
Business intelligence engineering
Demonstrates transformation of validated analytical data into decision-oriented semantic models and dashboards.
Dimensional data modeling
Demonstrates fact-and-dimension modeling for consistent healthcare analysis.
Analytics validation
Demonstrates explicit validation of dataset dimensions, KPI outputs, and metric interpretation.