Skin diseases represent a significant global health concern due to their increasing prevalence and the limited availability of dermatology specialists, particularly in resource- constrained regions. Early and accurate diagnosis plays a crucial role in improving treatment outcomes and reducing disease progression. This paper presents an AI-Based Skin Disease Diagnostic System that leverages deep learning and transfer learning techniques for automated skin disease classification. The proposed system employs the MobileNetV2 architecture, chosen for its computational efficiency and strong performance on image classification tasks, making it suitable for real-world deployment. The model is trained and evaluated using the HAM10000 dataset, consisting of dermatoscopic images spanning seven different skin disease classes. Image preprocessing and data augmentation techniques are applied to improve generalization and robustness. The system achieves an overall classification accuracy of 78%, demonstrating its potential as a clinical decision-support tool. The proposed solution is implemented using TensorFlow and deployed via a Streamlit-based web interface for interactive usage. Ethical considerations are emphasized, and the system is explicitly designed as a supportive diagnostic aid, not a replacement for professional dermatologists.\\n\\n