Fong Jian Zhong · Johor, Malaysia
Diploma in Computer Science student at TAR UMT (CGPA 3.38), certified in AI by IBM, Cisco, and Google Cloud — from machine learning foundations to hands-on prompt engineering with Gemini. The Sudoku solver on the right is my actual coursework algorithm, running live.
The same best-first search + Minimum Remaining Values strategy I lead in my Fundamentals of AI group assignment — reimplemented in JavaScript for this page.
Verified on Credly
Skills verified across these badges include machine learning, deep learning, neural networks, NLP, computer vision, chatbots, AI ethics, prompt engineering, and Gemini APIs. Verify on Credly · Google Skills · Cisco Learning Network
Selected work
End-to-end reinforcement-learning fine-tuning of Llama 3.2 3B Instruct using TRL's GRPOTrainer — 4-bit NF4 quantization, LoRA adapters, a custom rule-based reward function, and W&B experiment tracking, all on a free Colab T4. Published to Hugging Face.
View on Hugging Face →A native Android floating chat assistant that overlays on YouTube. An AccessibilityService detects the current video's title, channel, and timestamp, then streams context-aware answers from the Gemini API in real time.
View on GitHub →Group leader for my Fundamentals of AI assignment. I built the Informed Search (A*) component in Python — heap-based best-first search with an MRV heuristic — and coordinated teammates covering DFS and Simulated Annealing.
Try the live demo above ↑Toolbox
Study
Best grades: Calculus & Algebra (A+), Probability & Statistics (A), Discrete Mathematics (A), Software Development Fundamentals (A). Currently studying Fundamentals of AI, Data Structures & Algorithms, C++, and Mobile App Development.
CGPA 3.38