Projects

Stock prediction app combining multiple ML techniques with real-time data and sentiment analysis.

• Built time-series prediction with Prophet, Random Forest, Gradient Boosting, and SVR.
• Streamlit UI for 1,000+ stocks with indicators (RSI, MACD, Bollinger, ADX, ATR).
• Applied VADER sentiment; tuned models to reach ~87% predictive accuracy and reduced MAE by 23%.

Real-time collaboration with HD screen sharing, video streaming, and live chat.

• React front-end with Node/Express back-end; Socket.io for real-time messaging.
• MongoDB for users/rooms; deployed back-end on AWS EC2 and front-end on Vercel.
• Served 400+ unique visitors and 100+ rooms created.

Fantasy football team manager with live NFL data and standings.

• Integrated SportsDataIO API and Axios for real-time scores and standings.
• Dynamic EJS pages for player stats and game scores; tabulated regular season standings.
• Stored user-selected players and data in MongoDB and PostgreSQL with validation.