<p>The rapid shift to online learning has increased the need for reliable and effective online proctoring systems to maintain academic integrity during remote exams. This paper presents the Three-Eyed Invigilator, a semi-automated proctoring solution that enhances exam monitoring using an interactive and controllable rotator designed and implemented to provide a 360° view of the student’s environment. The system collects data from multiple sources, including webcams, front and rear mobile cameras, and microphones, to detect various forms of cheating. By combining advanced AI models for face recognition, object detection, and voice activity detection, it addresses the limitations of existing solutions that often rely on restricted data streams and the untested application of AI models. The adaptable interactive design allows the system to operate on both local and cloud-based platforms, ensuring flexibility and scalability. Extensive experiments demonstrated the system’s high accuracy in detecting cheating across predefined scenarios, emphasizing its robustness and efficiency. In tests with students across six scripted cheating scenarios, certain model combinations achieved a 90% cheating detection rate while processing each frame in &#xa0;0.02 s, meeting real-time proctoring requirements. The results highlight the effectiveness of the designed interactive system in upholding academic integrity, providing insights for future improvements in online proctoring technologies.</p>