FraudSense AI — Fraud Risk Analytics
MISSION
End-to-end fraud detection pipeline: SQL pattern mining on 1.85M transactions → EDA → ML modeling → Power BI dashboard → AI explanation API.
STACK
UNIT SELECTION
◀TERMINALUNIT PROFILE · ACCESS LEVEL: OPEN · RECORDS: 7
PROJECT FILES · 07
↑↓ SELECT · ENTER → CODE
MISSION
End-to-end fraud detection pipeline: SQL pattern mining on 1.85M transactions → EDA → ML modeling → Power BI dashboard → AI explanation API.
STACK
MISSION
Credit default probability model on 150K+ records: full EDA, feature engineering, multiple ML model evaluation, and classification precision improved via decision threshold optimization.
STACK
MISSION
Interactive financial KPI dashboard on Adventure Works data: revenue, costs, margins, and commercial performance across product lines.
STACK
MISSION
Relational database design and analytical queries for a fast food ordering system — data modeling, multi-table JOINs, aggregation and window functions.
STACK
MISSION
Automated sales reporting and trend analysis built entirely in Google Sheets: live KPI dashboard, formulas, and dynamic charting.
STACK
MISSION
Inventory optimization solution: ETL/EDA in Python, 500K+ records centralized in BigQuery, Power BI dashboards for real-time support, and a Random Forest overstock risk model integrated into Streamlit.
STACK
MISSION
Geospatial analysis to determine optimal countries for biotech lab expansion using 2021 COVID-19 case data and vaccination correlation modeling.
STACK