Skip to content

Navigation Menu

Sign in
Appearance settings

Search code, repositories, users, issues, pull requests...

Provide feedback

We read every piece of feedback, and take your input very seriously.

Saved searches

Use saved searches to filter your results more quickly

Sign up
Appearance settings
@sokcho-kim
sokcho-kim
Follow

sokchokim sokcho-kim

🍓

Block or report sokcho-kim

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
sokcho-kim /README.md

Hi, I'm Jimin 👋

AI Engineer exploring the intersection of Healthcare, Law, and Biology. I build traceable agentic workflows, not just chatboxes.

Previously a Web Developer, now orchestrating Multi-Agent Systems to solve domain-specific problems.

🚀 Why I Build

I specialize in converting unstructured "messy" data (Medical guidelines, Legal codes, Protein structures) into computable & searchable knowledge graphs.

  • Current Focus: GraphRAG, Multi-Agent Orchestration (LangGraph), Evaluation Pipelines.
  • Next Move: Optimizing AI infrastructure with Rust & Deep diving into Bio-informatics (Protein Folding).

🛠 Tech Stack

  • Core: Python, PyTorch, FastAPI
  • Agent & Search: LangChain/LangGraph, Neo4j, Qdrant, OpenSearch
  • Learning: Rust (for high-performance serving), Flow Matching

⚡ Featured Projects

**🏥 MedClaim **

AI-Powered Medical Insurance Review Agent

Live Service | Featured in MedicalTimes | DBR Case Study

A commercial RAG solution addressing the complexity of Korean medical insurance standards.

  • The Challenge: Automating reviews based on constantly changing government notices (고시) and complex calculation rules.
  • My Role: Engineered the Domain-Specific RAG Pipeline handling messy regulation updates.
  • Key Tech:
    • Traceability: Answers include direct citations (source/clause) and calculation logic.
    • Hybrid Search: Orchestrated OpenSearch (Keyword) + Qdrant (Vector) for precise retrieval.
    • Impact: Presented at KHF 2025; Successfully deployed for field usage.

🧬 Protein Structure & Drug Discovery (Research Interest)

  • Exploring Flow Matching with Transformers to optimize protein folding models (beyond AlphaFold3).
  • Goal: High-throughput screening for drug discovery.

🎯 Why Hackathon?

I want to push the boundaries of Claude Code. I aim to build self-maintaining subagents that can autonomously track domain knowledge changes (like new laws or research papers) and update their own testing pipelines—automating the "ResearchOps" workflow.


LinkedIn

Pinned Loading

  1. Hyeongseob91/Wanted_BeMyMuse Hyeongseob91/Wanted_BeMyMuse Public

    KoGPT2 언어 모델을 활용한 사용자 맞춤형 감성 작사 모델 서비스 - 2025 팀 프로젝트

    Jupyter Notebook 2

  2. AstralLoom AstralLoom Public

    Spin starlight and craft your own world.

    Python

  3. Korean_AI_Kiosk_Agent Korean_AI_Kiosk_Agent Public

    Forked from Wanted-AI-project/Korean_AI_Kiosk_Agent

    Auto Kiosk Agent based STT and TTS (for Korean cafe)

    Jupyter Notebook

  4. Jacod97/DACON-construction-accident-prevention Jacod97/DACON-construction-accident-prevention Public

    Jupyter Notebook 3

  5. wonjeonghwan/Kiosk_vision wonjeonghwan/Kiosk_vision Public

    Jupyter Notebook

AltStyle によって変換されたページ (->オリジナル) /