The LocalBrain repository is a full-stack system that serves as a local “second brain” for AI agents and applications. It features a FastAPI backend, Electron-based frontend, and a connector framework that ingests data from sources like Gmail, Slack, and browser history into a local markdown vault. The system employs two specialized agents — one for ingestion and another for retrieval — and uses a custom MCP (Model Context Protocol) layer to manage and serve personalized context to AI models.

The NBA Sports betting repository is a Python module that leverages sports APIs and machine learning to predict player props for games and identify valuable betting opportunities. It integrates a PostgreSQL database for efficient data storage and management. Currently, the project utilizes a Random Forest Regression model to analyze historical data and optimize predictions, enhancing betting decision-making. Development is ongoing, with plans to implement an LSTM model to further improve forecasting accuracy.

AI-Snake-Agent is an advanced reinforcement learning project that trains an AI to play the classic Snake game. Using a Deep Q-Learning algorithm, the agent continuously interacts with the game environment to develop optimal strategies. The project is built with Python, utilizing PyTorch and Pygame for training and simulation. Over time, the AI improves its gameplay, achieving higher scores through learned decision-making.

Movie Library AI is an advanced web application that lets users explore and discover a vast collection of over 45,000 movies. Users can browse detailed movie information, including themes, genres, and reviews, and easily add their choices to a personalized library. Additionally, Movie Library AI provides custom movie recommendations based on the user's current library.

Designed and developed a personal portfolio website utilizing React Js and CSS, showcasing a collections of projects