Abstract: In the quest for a more tailored cinematic experience, this paper presents a Movie Recommender System that harnesses the power of Cosine Similarity within a machine learning framework. The system’s cornerstone is its ability to discern nuanced user preferences and suggest films that resonate on a personal level. Employing Python, the research delineates a methodology that encompasses data collection, preprocessing, and feature extraction from a comprehensive dataset of movies. The crux of the system lies in its application of the cosine similarity algorithm, which calculates the affinity between movies and users based on shared characteristics such as genre, director, plot, title and cast. This approach is rigorously evaluated using metrics like precision, recall, and accuracy, ensuring the recommender’s reliability. The experimental design bifurcates the dataset into training and testing subsets, fortifying the system’s robustness. The paper’s findings illuminate the system’s efficacy in delivering precise and individualized recommendations, thereby augmenting the user’s movie selection process. It underscores the potential of the system to revolutionize content discovery and consumption in the entertainment domain. The conclusion encapsulates the project’s ambition to refine the recommender system further by integrating diverse machine learning algorithms and scaling its capabilities.