Developed a stock hedging platform using linear regression, statistical analysis, and historical market data to forecast systematic risk, integrating machine learning models and real-time data scraping tools, enhancing risk management and achieving a 356.46% portfolio increase
Engineered and deployed an AI-powered mental health peer support platform at MHacks2024 Hackathon, utilizing MongoDB for scalable data storage, Streamlit for an interactive userinterface, and NLP techniques combined with K-Means clustering to match users based ontheir worries and emotional states; seamlessly integrated Magic Loops API for automatedmeeting scheduling, providing real-time peer support with empathetic, data-driven matching
Developed a full stack AI SaaS platform using Next.js, MongoDB, Stripe, and Cloudinary forimage transformations, with secure route protection, serverless functions optimized by cache-based webhooks, and user authentication to enhance efficiency and streamline operations
(Pytorch, Flask, NumPy, Pandas, Python, HTML, CSS) Developed a machine learning-based movie recommendation system using Matrix Factorization in PyTorch, enabling personalized movie suggestions based on user ratings.Created a web application with Flask to handle user interactions, allowing users to log in, add ratings, and receive real-time movie recommendations, while leveraging TMDB API for fetching movie details.
(LC2K Processor, LC2K Simulator, C, Visual Studio) Developed a high-performance C program for translating LC2K assembly files (*.as) into machine code representations (*.mc), automating the conversion process and enhancing efficiency in programming workflows.Created a C program to simulate the LC2K ISA using machine code files, providing detailed processor analysis and generating output to stdout.
(Node.js, Express.js, HTML, CSS, Javascript, PHP, MySql) Developed a full-stack website featuring advanced functionality such as password hashing, password reset, email confirmation, and user data storage using cookies and tokens.