Computer Science Student @ EMSI | Full-Stack Developer & AI Enthusiast

I’m a computer science student working across local AI deployment, self-hosted infrastructure, and front-end development.

Adsum is a comprehensive, web-based smart classroom attendance system designed to replace manual tracking with an automated, AI-driven architecture.

Key Features & Architecture:

• Real-Time Facial Recognition: Powered by InsightFace and OpenCV on a Python backend to detect and verify students with high accuracy, utilizing GPU acceleration when available.
• Interactive Web Dashboard: A responsive frontend interface featuring live MJPEG camera streaming with bounding boxes, real-time metric tracking (AI FPS, current student counts), and attendance log visualization.
• Robust Backend API: Built with Flask, exposing RESTful endpoints for deep camera controls, runtime detection resolution switching, and dynamic recognition modes (continuous vs. interval-based tracking).
• Smart Logging & Registration: Implements strict "one-attendance-per-student-per-day" logic to prevent duplicate database entries, alongside the ability to register new students directly from the live camera feed into the system's tracking database.

AI & Machine Learning

Frontend & Backend

Infrastructure & Environment

I run dedicated local setups for model training, testing, and self-hosted media/services: