This study presents Doctor AI, an innovative hybrid deep learning framework that combines CNNs and Transformers to advance the classification of ocular diseases. By harnessing the strengths of CNNs for localized feature extraction and Transformers for capturing intricate global dependencies, the model demonstrates exceptional diagnostic performance. Using the ODIR-5K dataset, this research addresses class imbalance through advanced data augmentation strategies, achieving a notable F1-Score of 94.89% and an AUC of 99.10%. Furthermore, the Doctor AI system is designed for deployment across web and mobile platforms, featuring intuitive interfaces and secure data management to improve accessibility, particularly for underserved populations. While challenges related to dataset generalizability persist, future efforts incorporating transfer learning and larger, more diverse datasets aim to enhance its applicability. This hybrid approach establishes a new benchmark in automated ocular disease diagnosis, contributing significantly to scalable and efficient telehealth solutions.