Core Projects
Airline Payment Fraud Decision Intelligence System
An end-to-end fraud decision intelligence system for airline payment transactions. It predicts transaction-level fraud risk using ML, explains the factors behind each prediction with XAI, and translates risk into actionable payment decisions. The system combines predictive modeling, explainability, and business-oriented risk recommendations in a deployed web application, creating a complete workflow from transaction data to risk-aware decision support.
GitHub
Customer Retention & Revenue Optimization System
Built an end-to-end data science system to identify high-value customers, predict churn and purchase behavior, and optimize targeting strategies to maximize revenue under budget constraints. The project focuses on turning predictions into actionable business decisions with measurable impact. It combines data engineering, modeling, and optimization into a structured workflow that reflects real-world decision-making.
GitHub
End-to-End Time Series Analysis & Forecasting with Nutrition Data
A fully documented Time Series Analysis project built on a realistic multi-year nutrition dataset. The project progresses from dataset design and proper time handling to trend, seasonality, and variability analysis, followed by simple, interpretable forecasting. It focuses on understanding real-world time-based behavior, avoiding common pitfalls, and communicating insights clearly through visual storytelling.
GitHub
Find me on: