Aditya's Hackathon projects

Developed a machine learning solution to optimize fantasy sports team selection by building a web scraper to collect player data from ESPN and designing an AI-driven algorithm using XGBoost and PuLP. Integrated player form analysis and multi-objective optimization to balance constraints (budget, roles) while maximizing points, enabling efficient team selection. Implemented MLOps pipelines for scalability, leveraging Python, Pandas, and Scikit-learn to deliver a robust, data-driven decision-making tool.

Developed an educational search engine that spits out neatly organized responses to user queries, making it easier for everyone to learn. When someone searches, the engine creates a custom summary based on their age and education level, highlighting the most important points. We use past search data along with user details to fine-tune these responses, which also feed into our book and study material recommendation system powered by a TfidfVectorizer ML algorithm. I moderated the project by keeping our four-member team’s worktree in top shape, managing devops operations, and working on both the backend and the UI. For more details, please visit the link above.

Developed a security surveillance platform leveraging CCTV networks and a custom ML model with attention layers for real-time fight detection. Integrated live video feeds, historical anomaly logging, and interactive analytics using Chart.js. Implemented role-based access control, voice/text broadcasts, and rate-limiting via Flask-Limiter. Designed a modular architecture with Flask blueprints and SQLite for scalable data management. Features include browser-based voice alerts, anomaly trend visualization, and automated data-clearing scripts, providing a comprehensive solution for public safety monitoring.