This is a chronic neurological illness. It causes irreversible intellectual loss, impairing remembrance, thinking, as well as the difficulties to perform everyday job. It is a main reason of dementia, distressing millions of people worldwide. With over 46.8 million individuals currently suffering from AD, early diagnosis is critical as there is no existing cure to reverse or halt its progression. This paper proposes an automated method for brain images and differentiates between healthy and Alzheimer’s-affected brains. The method focuses on multiclass classification to predict different stages of AD, addressing the challenge of extracting meaningful features from medical images like MRI. Recent advancements in machine learning and medical imaging, such as brain structural MRI scans provided new hope. Our study explores the optimization of pretrained Reset models for disease detection, MRI data to identify morphological changes in brain structures. By employing AI-driven techniques, this research aims to get the accuracy which is often hindered by a shortage of medical professionals. The proposed system provides an automated framework for classifying AD stages, which could help in early detection and monitoring disease progression. The outcome prove that the model can help healthcare professionals in diagnoses that reduces the burden on medical staff.