Public portfolio simulation based on real work at Fannie Mae

What This Demonstrates: End-to-end analytics engineering skills for governance use cases:

• Dimensional data modeling (star schema with fact and dimension tables)
• Modern data stack architecture using DuckDB as analytical database
• Interactive dashboard development with Streamlit
• Self-serve analytics for non-technical stakeholders

Technical Implementation:

• Data Layer: DuckDB database with dimensional model
  • Fact table: Model review events, risk scores, SLA metrics
  • Dimensions: Model metadata, risk categories, review status
• Transformation: SQL for aggregations, window functions, risk calculations
• Application: Streamlit dashboard with filters, drill-downs, alerts
• Dataset: Simulated governance metadata (200+ models)

Skills Showcased: ✓ Data modeling (dimensional design) ✓ SQL (complex queries, CTEs, window functions) ✓ Python (Pandas, data transformation) ✓ Modern data stack (DuckDB - columnar analytical database) ✓ Dashboard development (Streamlit) ✓ UX design for stakeholder enablement

Real-World Context: At Fannie Mae, I build similar governance analytics systems monitoring 1200+ models. This project demonstrates my approach to:

• Translating business requirements into data models
• Building self-serve analytics applications
• Enabling non-technical users to make data-driven decisions

Link: [Live dashboard / GitHub]