Skip to content
User avatar
GitHub profile
LinkedIn profile

Kian Lok Chin

GitHub profile
LinkedIn profile

@kianlok

Kian Lok Chin

@kianlok

Deep Learning

Petaling Jaya, Malaysia

👋 Hi, I'm Kian Lok Chin

High-achieving Computer Science student specializing in AI Engineering, Generative AI, and Full-Stack Architecture. Experienced in building production-grade RAG systems, multi-modal agents, and scalable automation workflows.

🚀 Key Technical Skills

  • Programming Languages: Python, Java, Kotlin, R, SQL, JavaScript, Dart

  • AI & Machine Learning: PyTorch, TensorFlow, SBERT, FAISS, Time Series Modeling (Prophet, LSTM, ARIMA), Computer Vision, YOLO, YAMNet, RAG

  • Frameworks & Web: Next.js, Nuxt.js, Flutter, MVVM Architecture, WebSockets, Framer Motion

  • Tools & Infrastructure: Git, GitHub, Docker, Traefik, Redis, PostgreSQL (pgvector), Supabase, n8n, Firebase, PowerBI, Matplotlib

  • Languages Spoken: English, Malay, Chinese, Cantonese

🌟 High-Impact Projects

Zestix AI Automation Platform | Next.js, n8n, PostgreSQL, Vector Databases

  • Architected a unified AI portal featuring real-time WebSocket monitoring and a custom dark-mode UI using Framer Motion.

  • Engineered a 3-tier RAG context-injection pipeline with Supabase (pgvector) for high-precision retrieval.

  • Built 55+ n8n workflows utilizing Puppeteer, Apify, and Firecrawl to automate SEO audits and lead generation.

  • Orchestrated production deployment on Linux VPS using Docker, Traefik, and Redis-based API rate limiting.

Safe Vision | Python, PyTorch, TensorFlow, Generative AI, Flutter

  • Built a real-time threat detection pipeline integrating pose estimation (YOLO) and audio event classification (521 classes via YAMNet).

  • Designed a "SmartFusion" engine powered by Google Gemini to analyze multimodal data streams and minimize false positive alerts.

  • Developed a cross-platform mobile application in Flutter showcasing live threats on a 2D Digital Twin map synced via Firestore.

PeleeSENet (Transparent Medical AI) | Python, TensorFlow, PyTorch, CNN

  • Designed a hybrid CNN combining PeleeNet's dense connectivity with SE-Net and CBAM attention modules for early skin cancer detection (<3.5M parameters).

  • Integrated Grad-CAM to generate visual explanations for medical professionals, establishing clinical trust.

  • Implemented an adaptive learning loop for continuous model performance refinement based on expert feedback.

Brand Intelligence & Forecasting Platform | NLP, SBERT, FAISS, Time Series

  • Built an end-to-end analytics suite utilizing SBERT embeddings, FAISS clustering, and TF-IDF topic modeling.

  • Developed an ensemble engagement forecasting pipeline combining Prophet, LSTM, and ARIMA models with SHAP interpretability.

NutriTrack | Kotlin, MVVM, LLM APIs

  • Created a mobile nutrition application for clinical patient monitoring featuring an integrated LLM-based diet recommendation assistant.

💼 Experience

  • AI Solutions Architect (Intern) | Orangesoft Sdn. Bhd. (Nov 2025 – Feb 2026)

  • IT Student Researcher | Monash University Malaysia (Nov 2024 – Feb 2025)

🏆 Honors & Awards

  • Top 10 Finalist – Monash L'Oréal Datathon

  • IT Dean's Achievement Award 2024 – Monash University Malaysia (Average 90%+ across all units)

  • Unit Award for MAT1841 – Monash University Malaysia (Highest mark achieved)

  • JPA LSPM Scholarship Recipient – Public Service Department of Malaysia