This study is dedicated to design and develop a novel smart door knob based on AI and ML embedded with Arduino with the intention of facilitating enhanced security. The door lock utilizes face identification as a primary access mechanism thus eliminating the necessity for conventional keys. The system relies on the latest methods in deep learning computer vision and employs sophisticated CNNs for face detection and recognition in order to permit access only to users on the authorized list. The fastest and most accurate identification processes of facial features are performed using the OpenCV and Dlib libraries, while the servo motor that controls the lock is driven by an Arduino. The lock is also managed by the embedded platform, which controls its position depending on the recognition of the user’s face. To increase safety and convenience for the user, the door is equipped with a web interface that allows the user to remotely control the lock, register, and manage users. Users also receive alerts as these are happening, which greatly enhances system security. The system was put through extensive testing in different lighting situations and by different users. The success rate for various trials reached 95%. The face unlock feature is completed in less than five seconds which makes it easier to use in homes and small offices where security, safety, and convenience are paramount. These future upgrades will include the prevention of deepfake attempts, aid with scalability through cloud storage, and blend with other smart systems in the home for better overall security.