Skin disease presents a major global health challenge, and yet it remains unaddressed. The usage of deep learning techniques and models in skin disease detection has great potential to revolutionize diagnosis of skin diseases. One of the key approaches is to construct a dataset of multiple skin diseases and do analysis on them using advanced AI technologies like Deep Learning Neural Networks. This research mainly focuses on developing a detection mechanism powered by AI to detect multiple types of skin diseases. In this research we would be leveraging ResNet-101, a deep learning model for image classification tasks. The main goal is to provide a solution that is accurate as well as accessible which would allow individuals specifically in rural areas to analyze skin conditions without having to pay huge clinical diagnosis fees. Furthermore, the tool can help researchers and dermatologists in early stage diagnosis and further analysis of skin diseases. A major challenge in developing such a model is handling inconsistent and noisy data, which degrades the model’s performance and overall accuracy. To address this, pre-processing techniques, noise reduction filters and data augmentation were applied. Doctors and researchers can also help in examining and analysing the accuracy of dataset to build and develop better models for accurate detection. With the ISIC dataset, our research in developing a skin disease detection model achieved an accuracy of 99.3% making it an efficient tool for patients, medical professionals and researchers. With the help of AI-driven diagnosis, this research enhances early detection, medical research and accessibility, contributing to improvement of overall healthcare outcomes.