Attendance tracking is a crucial aspect of educational institutions, corporate environments, and various other organizations, but traditional methods like manual sign-in sheets or biometric systems are often time consuming, error-prone, and lack efficiency. This project proposes a novel attendance management system leveraging machine learning techniques, specifically deep learning models for facial recognition, to automate and streamline the attendance tracking process. The system employs a multi-stage approach, beginning with collecting and preprocessing facial image data from participants to train a convolutional neural network (CNN) based facial recognition model capable of extracting and identifying facial features accurately. The trained model integrates with a secure database for storing facial embeddings associated with participant information, enabling efficient comparisons during recognition. During operation, the system captures live video feed, performs real-time facial recognition, and matches detected faces with stored embeddings, logging recognized participants in attendance records and providing administrators up-to-date, accurate data. Incorporating a user friendly interface, administrators can monitor attendance, manage participant information, and configure system settings. Prioritizing ease of use, scalability, and adaptability across environments, the system undergoes continuous testing, feedback incorporation, and iterative development to ensure reliability and accuracy. By leveraging deep learning and facial recognition technology, this project streamlines attendance tracking processes, enhances operational efficiency, and offers a secure, innovative solution for attendance management applicable to diverse domains.