I like deep neural nets. - AI Engineer with 4+ years of specialized AI/ML experience and 19+ years of IT infrastructure background.
I design and build RAG systems, LLM applications, and end-to-end AI pipelines β from data collection and model training to cloud deployment and monitoring.
- π 1st place β Upstage AI OCR Competition (2025.12)
- π SCOPUS-indexed international journal publication β GCN-based Fire Situation Recognition (2024)
- π Led government R&D AI projects (ETRI, IITP) as AI Project Lead
- π€ Built RAG/NLP systems deployed in real-world services (112/119 emergency inference, B2B matching platform)
- βοΈ Experienced in MLOps: MLflow, GitHub Actions, AWS EC2/S3, Docker-based CI/CD/CT pipelines
- Graph Convolutional Network κΈ°λ° νμ¬ μν©μΈμ λͺ¨λΈ μ°κ΅¬ κ°λ°
- 95% prediction accuracy achieved Β· SCOPUS κ΅μ μ λ λ±μ¬
- AI Project Lead Β· λ°μ΄ν° μμ§/μ μ²λ¦¬/λΌλ²¨λ§/λͺ¨λΈ νμ΅ μ κ³Όμ μν
- μμμ¦ κΈμ κ²μΆ(Text Detection) λͺ¨λΈ μ€κ³ λ° Polygon μ’ν κ²μΆ μ΅μ ν
- 3,600+ μ΄λ―Έμ§ λμ κ°κ±΄ν(Robust) κ²μΆ νμ΄νλΌμΈ μμ±
- Data augmentation + hyperparameter tuningμΌλ‘ μΌλ°ν μ±λ₯ κ·Ήλν
- κΈ΄κΈμ κ³ ν μ€νΈμμ κΈ΄κΈμ±(Code 0~4)Β·μ¬κ±΄μ ν μΆλ‘ NLP λͺ¨λΈ λ° API κ°λ°
- AI Project Lead Β· λ°μ΄ν° κ³Όνμ 3λͺ , AI κ°λ°μ 2λͺ ν 리λ
- νκ΅μ μν΅μ μ°κ΅¬μ MATLABμΌλ‘ μ μλ μ νΈμ²λ¦¬ μ ν리μΌμ΄μ μ μλμ° λ°μ€ν¬ν± μ ν리μΌμ΄μ μΌλ‘ κ°λ°νκΈ° μν΄ C++ μννΈμ¨μ΄λ‘ λ³νν¨
- AI Project Lead Β· λ°μ΄ν° κ³Όνμ 1λͺ , AI κ°λ°μ 3λͺ ν 리λ
- 400GB λ¬μμ νΉν λ°μ΄ν° ETL νμ΄νλΌμΈ μ€κ³ β 30,000건 κ³ μλ JSON ν보
- Multilingual BERT κΈ°λ° κΈ°μ λ§€μΉ λΆλ₯ λͺ¨λΈ κ°λ° Β· 90% classification accuracy
- 3λ¨κ³ ν¬λ‘μ€ λ§€ν ν μ΄λΈ(IPC β KSIC) κ³ μ, VPN κΈ°λ° ν¬λ‘€λ§ μμ§μΌλ‘ 100,000건 λ¬μμ κΈ°μ DB ν보
- νκ΅μ μν΅μ μ°κ΅¬μ YOLO λͺ¨λΈμ μ΄μ©ν CCTV λμμ λ―Έμμ°ΎκΈ° νλ‘μ νΈ
- AI Project Lead Β· λ°μ΄ν° κ³Όνμ 3λͺ , AI κ°λ°μ 2λͺ ν 리λ
- End-to-end RAG μλΉμ€: μ§μ μ€ν¬λνν LangChain κΈ°μ λ¬Έμ β ChromaDB/FAISS λ²‘ν° DB ꡬμΆ
- Top-k κ²μ μ νλ 30% κ°μ Β· Docker Compose CI/CD Β· μ€νΈλ¦¬λ° UI ꡬν
- κ΅ν κ΅ν΅λΆ λ§€λ§€ λ°μ΄ν° + μ§νμ² /λ²μ€ μ μ§ λ°μ΄ν° κ²°ν©, Feature Engineering
- XGBoost/LightGBM μμλΈ λͺ¨λΈ Β· RMSE κΈ°μ€ λ² μ΄μ€λΌμΈ λλΉ 35% μ€μ°¨ κ°μ
| Year | Title | Venue |
|---|---|---|
| 2024 | Recognition of Fire Situation Using Graph Convolutional Network Model | SCOPUS-indexed International Journal |
| 2023 | GCN λͺ¨λΈμ μ΄μ©ν νμ¬ μν©μΈμ | νκ΅μ 보μ²λ¦¬νν νμ λ°νλν |
| 2023 | Graph Convolutional Network λͺ¨λΈμ μ΄μ©ν νμ¬ μν©μΈμ | μμ¬νμ λ Όλ¬Έ |
| π₯ Upstage AI OCR Competition 1st Place | 2025.12 |
| π μ 보μ²λ¦¬κΈ°μ¬ | 2012 |
| π CCNA (λ§μ 1000/1000) | 2006 |
| π§ LPIC-1 | 2013 |
- π§ sizin@naver.com
- πΌ LinkedIn
- π GitHub


