Technologies such as data mining and data visualization improve healthcare in the modern world. Hospital wait times can be lengthy, and after a prescription is filled, there may be additional costs. This can lead to costly hospital expenses. The goal of this research is to develop a health prediction system that gathers medical information and symptoms, then uses that information to anticipate the disease. In addition, it provides disease prognoses, medical advice, and a platform that matches patients with qualified physicians based on proximity and experience. In order to analyze and find patterns and insights in large patient datasets, the study focuses on understanding data mining techniques, such as the K-Nearest Neighbor Method, Decision Tree Algorithm, Random Forest Classifier, Logistic Regression & Support Vector Machine Algorithm, as well as data visualization techniques. Algorithms are employed to help in diagnosing different illnesses such as heart disease, kidney disease, diabetes, and breast cancer. After preprocessing the datasets, the aforementioned algorithms are applied to various diseases, and evaluation metrics are calculated for each algorithm and disease. Based on the results, accuracy was measured to find the robust algorithm suitable for diagnosing most of the diseases. The findings revealed that the algorithm Random Forest achieved the highest accuracy among all algorithms for each of the diseases mentioned and it is clear from the results that the algorithm has demonstrated its reliability when compared to other algorithms, this further strengthens the effectiveness of the algorithm in diagnosing diseases within the Smart Healthcare System.