Abstract
In the modern world, artificial intelligence (AI) technologies are increasingly being integrated into the field of sports and fitness, creating new opportunities for personalizing the training process. This article presents the concept and implementation of a personalized fitness trainer based on AI, designed to automatically monitor the correctness of exercise execution. The trainer uses computer vision technologies such as MediaPipe and BlazePose to recognize user body movements in real time, assess exercise technique, and provide tailored recommendations. The software architecture is described, including stages of data collection, normalization, and analysis using pre-trained models. The effectiveness of the proposed solution is evaluated, demonstrating its potential to reduce injury risks and enhance user motivation through real-time visual and audio feedback. The developed system is intended for both professional athletes and fitness enthusiasts aiming to improve their physical performance. The introduction of such AI-based solutions contributes to the accessibility of high-quality training support and promotes the development of healthy lifestyle habits.