Summary
Hwan Heo is an AI-focused developer and researcher with 9 years of experience applying Generative AI, LLMs, and computer vision to real-world industrial problems. He designs end-to-end systems—from data pipelines and model fine-tuning (PEFT, LoRA, 4-bit quantization on Llama3.1-8B) to real-time edge inference and user-facing interfaces—and recently architected a GenAI-powered, multi-modal RAG business-plan automation platform using ChatGPT-4.1, OCR, web search and agentic workflows. His background spans vision AI projects (underwater FLOC segmentation, sewer defect segmentation, on-device counting), process optimization for water treatment and EMS, and production-scale platforms integrating Kafka, Hadoop and Elasticsearch. With a PhD in Electronics and Electrical Engineering and a strong publication/patent record, he blends rigorous research methods with hands-on deployment in constrained environments like Jetson and Colab Pro. Notably, he routinely tunes model hyperparameters directly in limited-resource setups (rank=16, lora_alpha=16 on NVIDIA L4) to maximize SFT performance while keeping systems production-ready.
9 years of coding experience
8 years of employment as a software developer
Bachelor's degree Computer Engineering, Bachelor's degree Computer Engineering at 학점은행 / National Institute for Lifelong Education
Associate of Engineering Dept. of Software Engineering (3-year program), Associate of Engineering Dept. of Software Engineering (3-year program) at 동양공업전문대학 / Dongyang Technical College
Doctor of Philosophy - PhD Electronics and Electrical Engineering, Doctor of Philosophy - PhD Electronics and Electrical Engineering at Dongguk University
Korean, English, Japanese