Overview
AI/ML Engineer Intern at CJ AI Center, focused on developing Multi-Agent LLM pipelines for Korean Drama script analysis using LangGraph framework.
Technologies
Key Achievements
- Multi-Agent Pipeline Development - Designed and implemented a multi-agent LLM pipeline using LangGraph to decompose complex tasks
- State-Based Workflow - Developed a state-based workflow system with conditional branching for parallel asynchronous execution
- Model Optimization - Optimized LLM output consistency through systematic prompt engineering and pipeline optimization
- Performance Enhancement - Implemented batching and asynchronous processing to enhance inference performance
Professional Development
- Collaboration - Led knowledge transfer (KT) sessions and established team conventions
- Project Management - Adopted metric-driven development approach with controlled experiments
- Technical Skills - Deep understanding of multi-agent systems and enterprise LLM deployment