With the rapid growth of digital communication, the problem of email spam has become a persistent issue, negatively impacting the user experience and information. This project focuses on leveraging machine learning techniques to effectively detect and filter out email spam, enhancing the efficiency and reliability of email communication. The project begins with a comprehensive preprocessing step that involves cleaning and transforming raw email data into a structured format suitable for analysis. Feature extraction techniques such as word embeddings convert textual content into numerical representations, capturing semantic meanings and contextual information. Machine learning algorithms, including Decision Tree, Random Forest, Extreme Gradient Boosting, and, Deep Learning algorithm including Long Short-Term Memory (LSTM) network are trained and evaluated on a labeled dataset of emails to identify the most suitable algorithm for spam detection. The evaluation metrics encompass accuracy, precision, recall, and F1-score, providing a comprehensive assessment of model performance. The project contributes to the larger goal of creating a seamless and secure digital communication experience by reducing the intrusion of spam emails.