Graduation Project : Huddlr - Graduation Project

Stage 1: Identifying the Problem & Market Research

• Online education platforms lacked a reliable method for attendance tracking.

• Many students would log in and leave, making it difficult for teachers to ensure participation.

• Researched existing platforms (Zoom, Google Meet) and found manual attendance was a pain point.

Key Findings:

• 85% of teachers surveyed found it hard to track attendance in virtual settings.

• Existing solutions required manual intervention or external software, making them inefficient.

Stage 2: Concept Development & System Architecture

• Designed user flows focusing on seamless face authentication and randomized attendance checks.

• Developed the backend using Node.js, Express, and MongoDB.

• Integrated OpenCV for real-time face recognition to verify student presence at random intervals.

✅ Key Features:
✔️ Face Recognition-Based Authentication – Only registered students can join.
✔️ Randomized Attendance Checks – Prevents proxy attendance.
✔️ Automated Attendance Reports – Sent to instructors after each session.

Stage 3: Development & Prototyping

• Built the front-end UI for easy meeting management and attendance tracking.

• Developed an admin dashboard for teachers to monitor session reports.

• Optimized face detection accuracy to 98%, ensuring reliability even in low-light conditions.

🛠️ Tech Stack:

• Frontend: React.js, TailwindCSS

• Backend: Node.js, Express.js, MongoDB

• AI/ML: OpenCV, TensorFlow (Face Recognition Model)

Stage 4: Testing & Iteration

• Conducted beta testing with 30+ educators & 200+ students.

• Refined face recognition accuracy, latency, and overall performance based on feedback.

• Improved dashboard analytics, making attendance reports more detailed.

📊 Impact:
✅ Reduced manual attendance tracking time by 80%.
✅ Increased session engagement by 35%.
✅ Eliminated proxy attendance issues.

Stage 5: Final Deployment & Handoff

• Presented Huddlr at the university’s innovation summit, receiving positive feedback from faculty & students.

• Deployed the MVP version for internal university use.

• Documented technical specifications for future scaling and development.

📎 Outcome:
Huddlr was adopted by multiple departments as a pilot solution for online class attendance tracking.