The tremendous growth in technology has called for ingeniously designed computer interfaces to persuade users to use digital devices in this modern technology era. Disabled people are frustrated with digital accessibility due to various reasons viz., no keyboard-only access, presence of small radio buttons or click boxes, and inaccessible captchas. One particular solution could be advancing digital inclusions for persons with physical disabilities like creating websites that are written in an easily accessible language, providing alternative versions of images and text, and ensuring that the content is available across all platforms and devices. The main issue in digital inclusion is the dearth of qualified and skilled staff who can offer tech support to challenging people. Moreover, traditional interfaces often struggle to provide an inclusive and efficient means of interaction for those with limited mobility. In this paper, the main focus is to propose advanced eye-tracking technology using AI to enable users to navigate websites effortlessly using recognized facial gestures. Approaches like one shot, two shot, CNN folding, CNN+ORB, and Normalized gaze vector are presented and the outcomes of these methods are compared. It can be observed that all CNN+ORB (Oriented FAST and Rotated BRIEF) outperform with a good accuracy of 71% and less offset in deviation from the detection point.