YU-CHI (Yuchi) KO
✉ r14946005@ntu.edu.tw ｜ GitHub ｜ LinkedIn
AI Systems & LLMOps · Agentic AI · Software Engineer
🎓 EDUCATION
National Taiwan University
M.S. in Data Science (College of EECS) Taipei, Taiwan | Sep 2025 – Present
- Lab: Machine Intelligence & Agentic System Lab (MSLAB)
- Advisor: Prof. Shou-de Lin
- Research Focus: LLM applications, intelligent trading systems, and agent systems
National Taiwan University
B.B.A. in Finance & Decisive Data Science (UIBP) Taipei, Taiwan | Sep 2021 – Jun 2025
- GPA: 4.04 / 4.3 (Total 192 credits)
💼 EXPERIENCE
Cathay United Bank
LLMOps Engineer Taipei, Taiwan | Jun 2024 – Dec 2024
- Production-Scale AI System: Co-developed and launched Cathay’s first internal AI system for regulatory queries; improved finance-domain correctness and raised retrieval accuracy from 70% to 94%.
- Advanced RAG & GraphRAG Pipeline: Implemented enterprise-grade RAG with preprocessing for financial & compliance documents; added entity-linking and multi-hop reasoning.
- Agent-based Workflow Integration: Introduced multi-agent orchestration — the company’s first production-ready agent system supporting memory, tool use, chart and slide generation. Built a security-aware malicious query detection module, improving safe refusal rate by 85%.
Genibuilder Inc.
AI Engineer Intern Taipei, Taiwan | Jul 2025 – Aug 2025
- Genetic Search + Agentic Optimization: Built an AlphaEvolve-inspired agent framework combining GA + LLM reasoning; achieved a 4.8% gap on 100-node TSP while reducing token usage by ~20x.
- Multi-Cloud Deployment: Deployed containerized LLM services across GCP and Azure.
Academia Sinica (IIS)
Undergraduate Research Assistant (NLP) Taipei, Taiwan | Jan 2024 – Jun 2024
- Financial Report Highlighting Dataset: Constructed a human-annotated financial report dataset for decision-oriented LLM evaluation.
Gorich (Financial AI Startup)
Prompt Engineer / Software Engineer Taipei, Taiwan | Aug 2023 – Feb 2024
- Early LLM Adoption (2023): Entered the LLM field extremely early (mid-2023) and built one of Taiwan’s first production-grade LangChain RAG systems before enterprise tooling matured.
🚀 PROJECTS
- GraphRAG & Retriever/Reranker Ablation Study: Evaluated retrievers and rerankers for financial and medical documents on accuracy, latency, and cost trade-offs.
- User Preference Prediction: Built a multi-category user intent classification system, outperforming baseline by 11%.
- LSTM + GARCH Volatility Forecasting: Designed hybrid LSTM+GARCH volatility models with improved stability under market regime shifts.
🏆 ADDITIONAL
- Awards:
- GenAI STAR National Competition: Top 10 / 700+ teams, Best Award in ELTV category.
- 2026 TSMC IT CareerHack: Finalist.
- Languages: Mandarin (Native), English (IELTS 7.5; TOEIC 895), Spanish (Intermediate).
- International Exchange: 6-month exchange program in Barcelona.
- Technical Leadership: NTU AI Club Technical Officer (NTUAI).
🛠 TECHNICAL SKILLS
- Core: Python, C++, SQL, Docker, Git, Linux, FastAPI, PostgreSQL, Redis, Neo4j, Cloud (GCP, Azure).
- LLM Systems: RAG, GraphRAG, LLM Agents, Tool Use & Memory, Fine-tuning, LangChain, LlamaIndex.