This research presents Quizly, a multimodal Android quiz application designed to enhance educational accessibility, particularly for differently abled users. The objective is to improve user interaction in quiz-based learning through the integration of machine learning and multimodal input methods. Quizly incorporates hand gesture recognition, voice navigation, and gaze tracking, leveraging Convolutional Neural Networks (CNNs) and a cross-platform tech stack including Python, React Native, MongoDB, Express, and Node.js. The system was developed using an incremental and iterative methodology, with usability testing conducted among 30 participants. Results indicate high user satisfaction and effective performance across input modes. The app achieved an overall input recognition accuracy of 89% across all participants. The findings highlight the value of user-centric design and responsive interaction in educational applications, offering a foundation for future innovations in inclusive learning technologies.