About Me

I am JH Kang, a Generative AI Engineer who enjoys working happily.

After obtaining a Master’s degree from the Graduate School of Artificial Intelligence, I have been conducting research and development in AI models. Since the release of ChatGPT, I have focused on becoming a Generative AI specialist, concentrating on developing services utilizing LLMs. I enjoy growing together with colleagues and strive to apply the latest research trends to practical products.

구현과 설계 기록

직접 확인한 동작과 설계 과정에서 달라진 판단을 글로 남깁니다.

전체 글 보기 →

Work Experience

Generative AI Engineer

SK AX (C&C) • 2022.01 - Present

Manufacturing AI Agent Platform (2025.08 - Present)

Developing an automated platform for creating and managing AI agents based on manufacturing expert knowledge.

Tax AI Service (2024.07 - 2025.07)

Developed a Multi-Agent based chatbot for tax law analysis, involving large-scale data preprocessing and hierarchical indexing.

Financial RAG Chatbot (2024.02 - 2024.07)

Implemented an Advanced RAG service for the financial domain, applying techniques like CoN, CoVe, and HyDE.

Text-to-SQL PoC (2024.01 - 2024.02)

Developed and served an sLLM for converting natural language to SQL queries for legacy DB access.

AI Coding with sLLM (2023.04 - 2023.12)

Fine-tuned and served a small LLM for IDE integration to assist with coding tasks.

Vision AI Engineer (2022.01 - 2023.05)

Developed object detection models for drone/CCTV imagery and industrial applications.

System PM

Mando • 2018.10 - 2019.07

Managed system development for vehicle braking systems.

Project Manager

Emerson Korea • 2016.12 - 2018.10

Managed EPC projects focusing on instrumentation.

Piping Engineer

Samsung C&T • 2012.01 - 2016.07

Designed piping systems for overseas projects (Dispatched to Australia).

Skills & Certificates

Programming & Frameworks

PythonLangChainLangGraphTransformersPyTorchTensorFlowPEFT

Certificates

  • Azure AI Document Intelligence Solution
  • Tensorflow Developer Certificate

Awards & Activities

  • SK Group AI Competition (2025.09) - 2nd Place (SLM Development)
  • LangChain Open Source Contributor (2025.08) - Fixed Azure OpenAI API bug (PR #32649)
  • SK C&C GenAI Idea Competition (2023.10) - 3rd Place
  • AI Grand Challenge (2020) - 3rd Place (Behavior Recognition Track)
  • SK Group Expert (DEVOCEAN) - Active Member (2024.02 - Present)
  • Internal Instructor & Mentor - GenAI Engineer Training (2024 - Present)

Education

Sungkyunkwan University

M.S. in Artificial Intelligence (2019.08 - 2022.03)

Sungkyunkwan University

B.S. in Mechanical Engineering (2006.03 - 2012.03)

Publications

First Author

  • Kang, Junhyung, et al. “A survey of deep learning-based object detection methods and datasets for overhead imagery.” IEEE Access 10 (2022).
  • Kang, Junhyung, and Simon S. Woo. “DLPNet: Dynamic Loss Parameter Network using Reinforcement Learning for Aerial Imagery Detection.” AIPR (2021).
  • Kang, Junhyung, et al. “Sarod: Efficient End-To-End Object Detection On SAR Images.” ICIP (2021).

Co-Author

  • Kim, Dongwoo, ... Kang, Junhyung, et al. “Deep Learning for Overseeing Indo-Pacific Bottlenose Dolphin Tourism Law Enforcement.” CAI Workshop on AI for Environmental Intelligence: The Past, The Present, and The Future (2025). [OpenReview]
  • An, Jaeju, et al. “VFP290k: A large-scale benchmark dataset for vision-based fallen person detection.” NeurIPS Datasets and Benchmarks Track (2021).
  • 전소원, 강준형, 황진희, 우사이먼성일. “국내 딥페이크 기술 현황 및 제도적 대응방안 연구” CISC-S (2020).

Thesis

  • M.S. Thesis. “Time series prediction with dynamic learning by reinforcement learning” (2022).