This study focuses on advancements of computer vision in the fitness sector, particularly in the field of exercise monitoring and pose correction. The main objective is to introduce technology into the healthcare and fitness space by creating a virtual assistant that functions similarly to a trainer but with a greater emphasis on identifying where corrections are needed rather than how they should be performed, which is typically provided by human trainers. Computer vision, a key branch of Artificial Intelligence, plays an important role in capturing frames from the given media. MediaPipe helps in detecting body landmarks. These landmarks are then processed to determine whether the user is performing the exercise correctly or not with the help of trained KNN and SVM models. Both achieved an average of 96% F1-score, which is a key metric of evaluation, which further enables the system to provide detailed feedback and track user progress over time. This work demonstrates the potential of intelligent fitness monitoring to address the gap between traditional coaching and technology-driven healthcare solutions, catering to the growing trend of stay-at-home fitness practices.