I am a AI developer working on master degree in SNU. I have experience in Federated Learning, ZKML, LLM, computer vision and backend engineering. I have worked on projects ranging from model quantization to Infernece optimization in Triton Inference Server. I am also an enthusiastic tutor and mentor, sharing my knowledge with students and peers.

• Federated Learning optimization and Model Privacy.

• Developed ensemble models for vision tasks and conducted quantization for YOLO models with Android embedding.
• Worked on Human Detection, Face Recognition, Emotion Recognition, Action Recognition, and Gaze Estimation tasks.
• Built and deployed a Retrieval-Augmented Generation (RAG) system using Triton Inference Server, Llama3, Milvus, and Redis.

• First place (CNN + AdamW, Modular Pipeline)

• President (Jan. 2024 ~ Aug. 2024)

• ZKML Project Developer
• Co-authoring the Paper: "A ZKML-Integrated CNN Framework for Secure and Efficient COVID-19 Pandemic Response"

• Flock.io Validator

• Awarded a consolation prize in the POSTECH Solar Power Generation Prediction Competition (Oct. 2023).
• Third place in a hackathon with 1,500 participants: Developed a project summarizing YouTube videos and related news articles (Aug. 2023).
  • Featured in a news article: 효자동 개발자

Feel free to reach out to me via GitHub or LinkedIn!

• https://www.linkedin.com/in/kun-hwa-lee-9414b1332/
• [email protected]

Thank you for visiting my GitHub profile! 🚀