The urgency of developing a mobile application to automate the collection, initial assessment, and transmission of data on kinematic parameters of patient body movements with Parkinson’s disease to doctors is due to the need to improve the quality of diagnosis and treatment for this condition. Goal. To develop a mobile application that records and objectifies symptoms of Parkinson’s disease by analyzing kinematic parameters of a patient’s movement using machine learning and computer vision techniques. Methods. When developing the application, we used the computer vision frameworks MediaPipe Object Detector and TensorFlow, which are pre-trained neural networks based on a large amount of data. These frameworks can accurately cope with the task in various situations, including different backgrounds, clothing, and shooting conditions. We also used a proprietary algorithm based on the use of a low-pass filter, window control, and Berg autoregression to recognize and evaluate tremor. Results. The CYPD mobile application (https://cypd.mobi/index_en.html) has been developed. It contains computer learning methods and neural networks that are trained to analyze large amounts of data. The application helps to collect information about a patient’s tremors and the results of motor and kinematic tests. Conclusion. The proposed technology for recognizing and analyzing human movements has the potential to be applied in other areas of medicine, such as neurology, as well as to solve other practical problems where human movement is studied.