Madhuja Mungase

Engineered a customer churn prediction model on Telco dataset using Random Forest & Decision Tree with comprehensive EDA and feature engineering.

• Achieved ~88% accuracy and 0.89 F1-score on 7K+ records using Random Forest & Decision Tree
• Applied EDA, feature engineering, label/one-hot encoding, scaling, and outlier detection
• Deployed interactive Streamlit app for real-time predictions with joblib model persistence

Interactive web application visualizing 6 sorting algorithms and 5 searching algorithms with real-time animations and educational features.

• Developed visualizer for 6 sorting (Bubble, Insertion, Selection, Merge, Quick, Tim) and 5 searching algorithms
• Implemented real-time animation with labeled states (Comparing, Swapping, Sorted, Found, Pivot)
• Followed MVC principles with clean UI and categorized dropdowns for enhanced user experience

Comprehensive Power BI dashboard for banking risk analysis and decision-making using multi-table relational data and advanced DAX functions.

• Built comprehensive dashboard improving loan approval strategies and client targeting
• Applied DAX functions (SUMX, DISTINCTCOUNT, SWITCH) to compute KPIs: $139.9M Loans, $111.49M Deposits
• Delivered interactive dashboards (Loan Analysis, Deposit Analysis, Summary View) with advanced visualizations

Professional-grade business website for a diamonds company featuring real-time email integration and fully responsive design across all devices.

• Integrated real-time email functionality using Email.js to receive customer inquiries instantly
• Built fully responsive UI ensuring optimal experience across mobile, tablet, and desktop devices
• Designed modern layout with smooth navigation and user-focused interactions for enhanced UX