The Personalized Movie Recommendation System is an end-to-end machine learning project designed to provide movie suggestions tailored to user preferences. The system uses content-based filtering, where each movie is represented by its descriptive metadata — including genres, cast, crew, keywords, and storyline overview. These textual attributes are transformed into numerical vectors using TF-IDF / CountVectorizer, allowing the application to measure similarity between movies using cosine similarity. When a user inputs a movie title, the system retrieves the closest matches and generates a list of personalized recommendations. A clean and responsive Flask web interface allows users to interact with the model, and the entire application is deployed on Heroku for online accessibility. This project demonstrates skills in NLP, feature engineering, similarity modeling, web development, and cloud deployment.