I build applied AI systems that work in real-world environments.

My focus is not just on models, but on how AI integrates into actual business processes — from data pipelines and system design to deployment and iteration.

I often start with messy, ambiguous problems and turn them into structured AI solutions that can be reliably used in production.

Before building anything, I prioritize understanding how people work — because the success of an AI system depends more on adoption than accuracy.

I’ve worked across Malaysia, Singapore, Thailand, Brazil, and the US, designing systems in environments where data is inconsistent and processes vary widely. These experiences shaped how I approach scalable and resilient AI.

Currently focused on behavioral data pipelines, LLM/RAG-based systems, and building AI that holds up beyond the prototype stage.

** material-category-mapping-ai** Auto-classification pipeline for ~100K multilingual material records. 80%+ accuracy within a standardized category schema. Human-in-the-loop design — accuracy compounds as users give feedback. Triplet Loss + Hard Negative Mining based training architecture.

** vendor-deduplication-ai** Dedup pipeline across ~50K global supplier records. No universal ID in overseas data, so tax ID used only as secondary signal. 3-stage hybrid: Blocking → ANN → Embedding similarity scoring. ~30% duplicate rate found across the dataset — data cleansing completed.

** Internal Policy Review RAG System** (internal project) Automated compliance and subcontracting law review system. LangChain + LLM + internal docs / external legal data. Local LLM setup using Ollama / DeepSeek.

** phishing-detection** AI proxy that answers unknown calls and detects scams in real time. Whisper STT + keyword scoring + GPT response strategy. Flask backend · GitHub Actions CI/CD · deployed on Render.

** kleague-analytics** Pass destination prediction from K League event sequences. LSTM vs Transformer comparison. Sequential behavior modeling.

Analytics / ML Python R SQL pandas scikit-learn PyTorch TensorFlow NLP / Embeddings Triplet Loss LangChain RAG Ollama Streamlit

Systems / Backend Flask Spring Vue.js Oracle DB GitHub Actions Render UiPath RPA

ERP / Enterprise SAP ERP SAP Ariba SAP SRM SAP BW SAP SAC