Deep learning systems are good identifying the traits obligatory to precisely comprehend multifaceted sequences. This work put up a deep learning-based attention GRU for a computerised method of identifying skin diseases. The portable computing device-compatible attention GRU paradigm has been demonstrated to be accurate and reliable. For accurate projections, the recommended approach effectively preserves stateful data. The system's efficiency has compared various cutting-edge models, including Fine-Tuned Neural Networks (FTNN), Convolutional Neural Networks (CNN), and Highly Deep Convolutional Networks for Large-Scale Image Identification developed by the Visual Geometry Group (VGG). On the HAM10000 dataset, the proposed strategy performs better than cutting-edge techniques with over 85% efficiency. It requires less processing effort because of its resilience in recognising sooner and with more severity than the affected area nearly two times fewer calculations than the traditional attention GRU approach. It helps individuals and dermatology identify the type illness using an image of the affected area in the early stages of skin sickness. These findings show that the recommended strategy can help medical practitioners identify skin abnormalities quickly and precisely, hence averting morbidity and future issues.