This research presents a cutting-edge approach to the biomechanical analysis and classification of weightlifting lifts, focusing on the snatch and clean and jerk techniques, using advanced deep learning and computer vision technologies. Our system, designed to enhance performance and prevent injuries, features an intuitive user interface allowing users to upload dual-angle videos of their lifts. By synchronizing frame capture for crucial positions, tracking barbell trajectory, and classifying the lifts’ success, the system offers comprehensive insights into weightlifting mechanics. Deep learning models, including ResNet50 and LSTM networks, were developed to analyze the captured frames and accurately classify lift outcomes. The system generates detailed PDF reports encompassing angle measurements, phase-specific remarks, and tailored recommendations for technique improvement. Enhanced usability features such as a user guide video and strict format validation ensure a seamless experience for athletes and coaches. The findings demonstrate high accuracy in lift classification, underscoring the potential of integrating deep learning with biomechanical analysis to advance sports performance. This research not only contributes significantly to the field of sports biomechanics but also provides practical tools for optimizing training and performance in weightlifting. The innovative application of computer vision and deep learning in this domain represents a major leap forward in sports science technology.