// Service
Tarek Abou Rjeily builds real-time AI exam-proctoring and exam-integrity systems for universities. His prototype is a computer-vision system that detects phones, head-turns, talking, and unknown faces during exams, logs photographic evidence per violation, and sends instant Telegram alerts to proctors — built on YOLOv8 and face recognition.
Cameras feed a computer-vision pipeline that watches for defined violations — a phone in frame, repeated head-turning, talking, or a face that doesn't match the enrolled student list. Each detection is logged with a photographic evidence frame and timestamp, and proctors get an instant alert, so intervention happens during the exam rather than in a dispute weeks later. This is the architecture of Tarek Abou Rjeily's prototype.
Yes — the same face-recognition pipeline runs in auto mode against the class timetable, marking attendance as students appear on camera. Universities get exam integrity and attendance automation from one system.
Typically one exam hall, one camera setup, and one real exam session under supervision, so the integrity office can judge detection quality on real footage before any wider rollout. Tarek scopes pilots in a short call with the university's IT and exam-integrity teams.
Message me on WhatsApp, email tar@tarekrjeily.com, or see all services.