Portfolio
Work Experience
AI Intern
Sep 2025 — Feb 2026Dynamic Traffic System Sdn Bhd
Rawang, Selangor
- Assisted with data collection and storage using Node-RED and pgAdmin for an STGNN-LSTM dataset
- Supported development of an STGNN-LSTM engine to optimize green time and reduce traffic congestion
- Manually labelled 15,000 car license plate images for ANPR dataset
- Trained YOLOv8 and YOLO-NAS models for license plate recognition and vehicle classification
- Performed manual vehicle classification to benchmark against YOLO inference results
- Refactoring ~10,000 lines of legacy Python and HTML/CSS to Next.js (in progress)
Projects
lokasibus.xyz
Real-time Rapid Bus KL tracker using WebSocket for live location updates. Built to provide commuters with accurate, up-to-the-minute bus positioning.
UITMGetTable – Timetable Generator
Next.js web platform that automatically generates UiTM student timetables. Integrated custom parsing algorithm and UI for easy course selection. Achieved over 1,000 website visits and 750+ timetables generated.
DISK Voting System
Secure online voting system used by 375 students. Developed RESTful APIs, managed user authentication, and created admin/user interfaces. Coded the entire backend in three days, ensuring secure authentication and data integrity.
ReLearn – Student Book Marketplace
Educational platform with separate admin and user modules. Implemented user registration, feedback, and resource management features for buying and selling used books.
Face Recognition System
Real-time face recognition system using DeepFace and OpenCV. Implemented facial detection, verification, and emotion analysis. Integrated webcam streaming for live recognition and database matching. Explored model performance with VGG-Face, Facenet, and DeepID.
PDF Summarizer (sykrOpenAI)
Uploads academic PDFs and summarizes content using OpenAI API for quick content extraction.