Dermatology remains a complex field due to the multitude of skin diseases and the uncertainties surrounding their diagnosis. These diseases exhibit variations influenced by environmental, geographical, and genetic factors. The human skin, with its diverse characteristics such as hair presence and variations in tone, further complicates diagnosis. Current skin disease diagnosis often involves a series of pathological laboratory tests to accurately identify the condition. Skin diseases can be severe, even fatal if left untreated, underscoring the importance of early detection and intervention. The proposed convolutional neural network (CNN) system aims to address this challenge by identifying eight common skin diseases: actinic keratosis, acne and rosacea, bullous disease, carcinoma and malignant lesions, bacterial infections, cellulitis impetigo, and others. The framework employs the “20 Skin Infections Dataset” from Kaggle, a dataset with shifting numbers of pictures per illness lesson. Whereas a few infections are overrepresented in this dataset than others, the CNN framework has been shown to have the potential for successfully diagnosing skin illnesses and hence being valuable in giving opportune and exact restorative medicines.