Seongjin Kim

MS Student · 2024.03 —

About Me

I'm a master's student in Industrial Engineering at UNIST, focusing on anomaly detection, and domain-specific RAG applications in real problems.

Research Interests

  • Retrieval-Augmented Generation (RAG) with domain knowledge and prompt engineering
  • Industry-academia collaboration in manufacturing, maritime, and finance
  • Robust anomaly detection for multivariate time-series under imbalance and noise

Milestones

  1. 2018.03 – 2024.02
    B.S. in Industrial Engineering

    Ulsan National Institute of Science and Technology (UNIST)
    Undergraduate research, student council, and mentoring activities

  2. 2024.03
    M.S. in Industrial Engineering

    Ulsan National Institute of Science and Technology (UNIST)
    Research on anomaly detection and maritime RAG systems

  3. 2024.03 – 2024.08
    LG Electronics · LQC Project

    Developed anomaly detection models for multivariate time-series sensor data
    Addressed data imbalance and label noise with domain-knowledge filtering

  4. 2025.01 – 2025.06
    SCSC Center Project · Human-Centered Carbon-Neutral Supply Chain

    Maritime fuel consumption prediction with environmental resistance modeling
    Multimodal fusion of AIS and weather data for sustainable shipping

  5. 2025.05 – 2025.09
    Maritime Chatbot System

    Built RAG framework using LangChain to answer AIS data and maritime regulation queries
    Designed hybrid prompting strategies and evaluation pipeline

  6. 2025.09
    RCA RAG for Maritime Accident Analysis (planned)

    Building RAG-based maritime accident root cause analysis system

Current Work

  • Maritime RAG system for AIS-based question answering
  • Maritime Root Cause Analysis by RAG system

Notes for Collaboration

  1. Interested in interdisciplinary projects at the intersection of AI and domain expertise
  2. Open to collaboration on maritime, finance, and manufacturing applications of AI
  3. Actively exploring multimodal AI and communication