Breast cancer remains a significant global health challenge, where early detection and appropriate treatment are crucial for improving survival rates. This study proposes an AI-driven system for breast cancer classification and personalized treatment recommendations, designed to assist clinicians in making accurate and informed decisions. Using advanced deep learning techniques, the system classifies breast cancer into stages (Stage 0 to IV) by analyzing mammogram images and clinical features such as tumor size, lymph node involvement, and hormone receptor status. Among the deep learning models tested, DenseNet-121 achieved the highest average accuracy of 98.88% demonstrating its superior performance in effectively classifying breast cancer images. Among the evaluated machine learning models, LightGBM demonstrated the highest average accuracy of 96.60% making it the top-performing model in this analysis. The system extends beyond classification by integrating a recommendation engine that provides tailored, stage-specific treatment suggestions, including surgery, chemotherapy, radiation, hormonal therapy, or lifestyle adjustments, based on established medical guidelines. This approach enhances diagnostic accuracy and bridges the gap between AI systems and clinical practice, empowering healthcare professionals with a reliable decision-support tool. The study aims to revolutionize breast cancer management by combining precise classification with actionable treatment insights, ultimately improving patient outcomes and care quality.