Fast advancing eye-tracking technology lets people run digital gadgets with only their eyes. Following a user’s eye and figuring their goals has improved accessibility, gaming, healthcare, and automobile safety. For specially abled individuals, keyboards and mouse might be trouble. Eye-tracking replaces and simplifies the requirement for hand-operated computer interface. This work aims to build an accurate, real-time, reasonably priced eye-tracking system using state-of- the-art computer vision and deep learning methods. The method seeks to communicate with many apps, decrease processing load, and improve gaze estimate accuracy. Deep learning, particularly using CNNs, has lately improved pupil detection and gaze estimate. These advances enable screen coordinate alignment and real-time gaze point tracking. Eye-tracking technology has difficulty with lighting, head movement, and eye shapes. While precise, traditional infrared tracking devices are costly and hard to get. We investigate how highly accurate conventional cameras and deep learning models may address these problems. Using OpenCV, Dlib, and TensorFlow, the proposed system detects gaze direction, eye areas, and face landmarks. Beyond accessibility, eye-tracking has uses. Gaze-based controls improve game immersion and involvement. Medical experts discover neurological conditions like Parkinson’s disease, autism, and attention deficit disorder using eye-tracking. It is used in vehicle safety to evaluate driver awareness and stop collisions. Eye-tracking monitors user behavior to help advertisers and website designers position ads and change their layout. This work presents a comprehensive method for creating a real-time eye-tracking system. The system is taught on a large dataset to accommodate ambient elements, illumination, and user demographics. Using grayscale conversion, Gaussian filters, and edge detection, the method increases pupil detection.