This article dives into the utilization of machine learning (ML) and deep learning (DL) algorithms for the diagnosis of glaucoma and diabetic retinopathy, significant ocular conditions. Glaucoma, marked by optic nerve deterioration, often stems from elevated intraocular pressure, while diabetic retinopathy results in damage to retinal blood vessels, causing vision impairments. The diagnostic process encompasses stages like data acquisition, preprocessing, data augmentation, feature extraction, and model selection. A diverse range of models, including convolutional neural networks (CNNs) and traditional ML algorithms, are deployed for disease classification. Evaluation metrics like AUC, Accuracy, F1 score, and DiceOC assess model performance. The article also addresses challenges such as data quality and ethical considerations, emphasizing the imperative of data privacy in healthcare AI research. Ultimately, this research endeavors to contribute to healthcare advancement by developing precise and efficient diagnostic tools for glaucoma and diabetic retinopathy.