I built a hybrid Spotify recommendation system that combines content-based and collaborative filtering. I developed its end-to-end ML pipeline and integrated recommendations into a Streamlit application.
For my YouTube Comment Analyzer, I built a sentiment analysis system using NLTK, TF-IDF, and LightGBM, then connected the model to a Flask API and Chrome extension. I used MLflow and DVC to track experiments and data, and deployed the Dockerized application on AWS EC2.
My projects also include a Swiggy delivery-time prediction system and a rental analytics application with price prediction, property recommendations, and market insights. I’ve worked with Python, machine learning, and deployment tools including AWS, Docker, and GitHub Actions.
