The advent of Artificial Intelligence (AI) and Machine Learning (ML) has significantly transformed the healthcare industry, especially in rehabilitation. PhysioAI is an innovative AI-driven physiotherapy support system that enhances home-based rehabilitation through real-time motion analysis using an ordinary device camera. The system integrates pose estimation algorithms (using OpenCV and CNN-based models) to detect body joints, evaluate posture, and provide immediate corrective feedback—without requiring wearable sensors. A prototype was tested with a group of voluntary participants performing basic physiotherapy exercises, where the system achieved an accuracy of over 90% in pose detection and received positive feedback for usability. PhysioAI promotes patient engagement, ensures adherence to prescribed routines, and aims to reduce the dependency on in-person therapy sessions. Future versions will incorporate escalation protocols and real-world clinical testing to further validate its effectiveness.