MyungHwan Hong — profile photo

MyungHwan Hong

AI Engineer

I build AI agent systems, develop RAG architectures, fine-tune LLMs, and curate datasets for reliable AI model training.

End-to-end production AI systems

About Me

Multi-agent systems, RAG architectures, and fine-tuned LLMs focused on automated decision-making in real-world applications.

Hi, I’m MyungHwan,

An AI engineer based in South Korea, specializing in LLM systems, multi-agent architectures, and Retrieval Augmented Generation.

My work focuses on building practical AI systems, including domain specific model fine-tuning, dataset engineering, and AI applications that automate complex decision workflows.

AREAS of FOCUS:

  • AI Agent Systems: Designing multi-agent architectures that coordinate reasoning, tool calling, and task execution using frameworks such as LangGraph and LangChain.
  • RAG Architectures: Building retrieval pipelines with vector databases, reranking strategies, and context optimization for grounded LLM responses.
  • LLM Fine-Tuning: Creating domain-specific model adaptation using LoRA / QLoRA and curated datasets to improve task performance and reliability.
  • Dataset Engineering: Constructing training datasets through data curation, feature extraction, and structured prompt generation for model development.
  • AI Applications: Developing practical AI systems including multimodal assistants, automation agents, and decision-support tools.
  • LLM Tool Integration: Connecting language models to external APIs, services, and automation workflows

Featured Projects

AI systems built with retrieval pipelines, tool integration, and fine-tuned models.

Tech Stack

These below are the core stack that I mainly use.

Core Stack

Python
Hugging Face Transformers
LangGraph
LangChain
PEFT
PyTorch
Pydantic
ChromaDB
NumPy / Pandas
Git
scikit-learn(basic usage)
Plotly
Autogen
Gradio

Contact

Open to amazing AI engineering opportunities! Please feel free to reach out via email at mh.hong@proton.me (or simply, LinkedIn).