Diabetes is a chronic illness with the potential to induce a worldwide health care crisis. As per the International Diabetes Federation, 382 million people exist with diabetes in the world currently. By 2035, the number is expected to be doubled as 592 million. Diabetes mellitus or diabetes is a disease generally induced due to the grown level of glucose in the blood. Physical and chemical tests are different traditional methods for diagnosing diabetes. However, diabetes prediction in advance is pretty challenging for medical practitioners due to complicated interdependence on multiple factors as diabetes influences individual organs such as eye, heart, kidney, eye, nerves, foot, etc. Data science techniques have the potential to help the medical field by answering some of the general questions. One such task is to facilitate predictions on medical data. Machine learning is the most useful technology for the medical field in data science. Machine learning is helpful because of the way machines learn from experience. This project aims to propose a valuable technique for earlier detection of the diabetes disease for a patient with higher efficiency by combining the outcomes of different machine learning techniques, the supervised machine learning methods including Decision Tree, Logistic Regression, Random Forest, Neural Network, XGBoost, and Support Vector Machine