Projects Portfolio

×

Black-Scholes Options Pricer - Technical Details

Project Overview

Comprehensive options pricing application implementing the Black-Scholes-Merton model for European options with real-time Greeks calculation and Monte Carlo simulation capabilities.

Key Features

• Black-Scholes Pricing: Analytical solution for European call and put options
• Greeks Calculator: Real-time Delta, Gamma, Theta, Vega, and Rho
• Volatility Analysis: Implied volatility using Newton-Raphson method
• Monte Carlo Simulation: 10,000+ path simulations for exotic options
• Volatility Smile: Interactive visualization of volatility skew across strikes
• Sensitivity Analysis: Heatmaps showing option value changes

Technical Stack

• Python 3.9 with NumPy, SciPy
• Streamlit for interactive web interface
• Plotly for visualizations
• Pandas for data handling

Code Sample

Results & Impact

• 99.9% accuracy compared to Bloomberg terminal prices
• Greeks validated against professional trading platforms
• Monte Carlo convergence within 0.5% of analytical solutions
• Processing time < 0.1 seconds for single option pricing

×

Tableau Dashboards - Technical Details

Project Overview

Interactive business intelligence dashboard analyzing Global Superstore sales performance with drill-down capabilities and geographic mapping.

Dashboard Features

• Sales and profit trends over time
• Geographic heat map of regional performance
• Product category profitability analysis
• Customer segment comparison
• Top/bottom performers identification
• Interactive filters and parameters

Insights Delivered

• Identified underperforming product categories
• Regional sales patterns and opportunities
• Customer segment profitability analysis
• Seasonal trends and forecasting

×

Juicee!! Sales Tracking App - Technical Details

Project Overview

Full-stack web application built for the Infinite Reach direct sales team to track daily sales performance, manage team members, and drive accountability through real-time leaderboards and detailed audit trails. Deployed on Netlify with a Supabase PostgreSQL backend.

Key Features

• Role-Based Access Control: Admin, Manager, and Rep roles with different permissions and views
• Real-Time Leaderboard: Live rankings updated instantly as sales are logged
• Sales Entry & Tracking: Reps log daily sales; managers review and approve entries
• Audit Trail: Full history of all changes with timestamps and user attribution
• Excel Export: One-click download of team performance data for reporting
• Team Management: Admin can add/remove team members and assign roles

Technical Stack

• Frontend: Vanilla JavaScript, HTML5, CSS3
• Backend: Supabase (PostgreSQL) with Row-Level Security policies
• Authentication: Supabase Auth with JWT tokens
• Hosting: Netlify with continuous deployment from GitHub
• Export: SheetJS library for Excel file generation

Business Impact

• Replaced manual WhatsApp-based tracking with a structured digital system
• Real-time visibility into individual and team performance
• Reduced reporting time from hours to seconds with automated Excel export
• Improved accountability through immutable audit trails
• Scalable across multiple sales teams and regions

×

Loan Default Prediction - Technical Details

Project Overview

Machine learning model predicting loan default risk using ensemble techniques. Analyzes borrower characteristics, financial history, and loan attributes to provide probability scores.

Key Features

• Feature Engineering: 50+ features including debt-to-income, credit utilization, payment history
• Ensemble Modeling: XGBoost, Random Forest, Logistic Regression with weighted voting
• Class Imbalance: SMOTE for synthetic minority oversampling
• Model Interpretability: SHAP values for feature importance
• Risk Segmentation: Low, Medium, High, Very High risk categorization
• Production Pipeline: Automated end-to-end prediction workflow

Model Performance

• Accuracy: 87%
• ROC-AUC Score: 0.91
• Precision: 83%
• Recall: 79%
• Dataset: 250,000+ loan applications

Top Predictive Features

• Credit Score (FICO): 25% feature importance
• Debt-to-Income Ratio: 18% importance
• Employment Length: 12% importance
• Loan Amount: 10% importance
• Credit Inquiries: 8% importance

Business Impact

• Identifies 79% of defaults, reducing bad debt exposure
• Enables approval of creditworthy borderline applicants
• Real-time predictions reduce processing time from days to minutes
• Explainable AI ensures transparency and regulatory compliance

×

Portfolio Optimization - Technical Details

Project Overview

Interactive Shiny dashboard implementing Modern Portfolio Theory using quadratic programming for optimal asset allocation.

Key Features

• Efficient Frontier: Visualization of risk-return tradeoffs
• Sharpe Ratio Optimization: Maximum risk-adjusted returns
• Minimum Variance Portfolio: Lowest-risk allocation
• Custom Constraints: Position limits, sector constraints, long/short
• Interactive Dashboard: Real-time portfolio adjustments
• Historical Backtesting: Performance validation

Results

• 15% annual outperformance vs equal-weight benchmark
• 25% reduction in portfolio volatility
• Real-time Yahoo Finance data integration
• Support for 20+ asset portfolios