Optical coherence tomography (OCT) is a non-invasive imaging technique with extensive applications in ophthalmology. This paper presents OpticIntel, a real-time system that classifies retinal diseases from OCT images using deep learning. The model is based on the VGG-19 architecture and was trained on a publicly available OCT dataset with over 2,000 labeled images spanning seven classes: AMD, DME, ERM, RAO, RVO, VID, and Normal. Using transfer learning, augmentation, and optimization techniques, the model achieved an accuracy of 90.88%, an F1-score of 89.61%, and an AUC-ROC of 96.90%. The system was deployed as a web and mobile application to enable scalable screening of retinal conditions. Comparative evaluation with VGG16 shows the superior performance of VGG19. The study aims to support ophthalmologists with an accessible AI tool that enhances screening and diagnostic accuracy.