About Me
A brief introduction
Ph.D. Student, Smart Design Lab
Jangseop Park (박장섭) is a Ph.D. Candidate at the Smart Design Lab, KAIST, under the supervision of Prof. Namwoo Kang. He previously served as an AI Researcher at NARNIA LABS, where he specialized in revolutionizing industrial product development through generative AI.
His current research explores the intersection of Large Language Models (LLMs) and Mechanical Engineering, with a focus on LLM for Engineering Applications and Data-driven Design Optimization.
To realize this vision, he develops Autonomous Design Systems that integrate AI agent systems and 3D deep learning to solve distributed engineering problems. He is advancing the field through frameworks like AutoATC, which automates complex multidisciplinary design decompositions, and by implementing domain-specific in-context learning pipelines for sophisticated optimization tasks such as LLM-guided adaptive penalty updates.
Simultaneously, he creates high-fidelity data-driven surrogate models using advanced techniques like Point-DeepONet, BMO-GNN, and Physics-constrained GNNs to predict nonlinear physical behaviors on non-parametric 3D geometries. His work bridges cutting-edge AI with traditional engineering workflows to establish fully autonomous optimization ecosystems.
Keywords: LLM for Engineering Applications · Multi-Agent Systems (MAS) · 3D Deep Learning · Surrogate Modeling · Data-driven Design Optimization
- Ph.D. Student · CCS Graduate School of Mobility, KAIST
- Advisor: Prof. Namwoo Kang (Smart Design Lab)
- Daejeon, South Korea
- jangseop@kaist.ac.kr