AI/ML Engineer and Researcher with a Master’s degree and 2+ years of production software engineering experience. I specialize in bridging deep learning research with end-to-end production AI platforms.
Key Highlights: • Production AI & LLM Systems: Built Bé Đậu (AI career prep platform featuring 3-stage LLM pipelines, strict Pydantic JSON schemas, local llama.cpp/Qwen inference, and Web Speech API voice mock interviews) and DichPDF (layout-preserving document translator). • Deep Learning Research: Developed state-of-the-art Few-Shot and Open-Set GNN models for CAN bus anomaly detection at Soonchunhyang University, achieving a 95% detection rate on unseen attack classes. Published in a Q1 journal (Knowledge-Based Systems). • Software Engineering at Scale: Shipped 15+ features for Zalo PC, optimizing local message search latency by 30% across large-scale TypeScript/Electron codebases.
Core Stack: PyTorch, PyTorch Geometric, LLMs, RAG, FastAPI, Next.js, TypeScript, Docker, GNNs, FAISS.
South Korea