doyoon lee

I’m a Native Korean speaker and AI / Korean Language Specialist focused on training, evaluation, and Korean NLP projects. In my daily creative and analytical workflows, I use large language models as a core practice—bringing both strong written Korean across registers and bilingual sensitivity (KO↔EN) to every task.

My core strength is practical, repeatable LLM use: I’m a daily LLM power user who works directly with Claude (Sonnet / Opus), ChatGPT, and Gemini, applying prompt engineering and structured output design to multi-step goals. I also do routine evaluation of model outputs in both Korean and English, so I’m familiar with where today’s models succeed and fail specifically in Korean.

Since 2024, I’ve been building personalized AI agent setups as an AI-Assisted Workflow Designer, using Claude Code, custom skills, and MCP servers for daily creative and research workflows. I maintain and iterate on pectrono on GitHub for personal tools and experiments, pairing self-directed learning with hands-on testing and editorial judgment.

I’m comfortable across casual speech, business prose, creative writing, and technical translation, and I have an editorial eye for what reads naturally in Korean versus translated-sounding output. If you need someone who can combine native-level Korean nuance with structured LLM workflows and model output evaluation, I’m ready to contribute.