cardiac illness covers several cardiac issues. According to WHO statistics, cardiovascular illnesses kill 17.9 million people worldwide. This worrying rise in heart disease cases shows how important it is to find it early. The health of the patient is closely linked to early and accurate diagnosis, so this is a very important issue. To deal with this, programmes that use machine learning have been made to find and identify people with heart disease. In this project, we made a heart disease prediction system that uses a patient's medical background to figure out how likely it is that they will be diagnosed with heart disease. The method looks at things like chest pain, high blood pressure, heart attack, high cholesterol, and high blood pressure. Anaconda Python software has been used for machine learning algorithm implementation, experimental data processing, and visualisation. This project demonstrates how Python and machine learning can detect cardiac problems. This project used UCI data. It has 3000 cases, and each one has 14 different characteristics. These examples show the results of different tests done to predict how well heart disease can be diagnosed. The dataset has been split into two parts so that the success of the algorithms can be judged. Seventy percent of the data will be used for training, and the other thirty percent will be used for tests. The study uses SVM, KNN, LR, RF, and DT algorithms to enhance heart disease detection. The goal of these programmes is to be able to tell if a person has heart disease or not. With this study, we want to compare how well different models work and look at the results to improve how well heart disease can be found. Using machine learning techniques and Python code together could be a good way to deal with this world health problem. The accuracy is calculated and analysed using these values. It has 14 classification attributes. We found that random forest is performing well out of all five algorithms giving an accuracy of 98.05%, Decision tree is performing next to random forest giving an accuracy of 97.08%, Support vector machine is giving accuracy of 90.25%, K-nearest neighbour and logistics regression are giving accuracy of 81.82%.