Training patient preventive approaches, the research involved decision trees and support vector machine (SVM) algorithms in a medical chatbot architecture. The demand for healthcare solutions is always growing because of the rise of chatbots as one of the most dependable means that users can have initial diagnoses and medical advice. In this study, the author has used natural language processing (NLP) to predict diseases using symptoms provided by patients and recommend appropriate medications for them, to compare the predictive capabilities of decision trees and SVM algorithms against each other. When it comes to medical chatbots, the suggested SVM algorithm outperforms decision trees both in disease prediction and medication recommendation. A machine learning-based system that utilizes advanced methods such as SVM enhances the chatbot’s diagnostic ability thus enabling better diagnosis while also giving timely recommendations on drugs. This study represents a contribution to AI-driven healthcare solutions aimed at improving patient outcomes while reducing the costs of healthcare. Decision trees, SVM, and KNN models are used in the research with an accuracy of around 97% at the time of model development.