Jihoon Kim

jihoon.kim@kaist.ac.kr · jihoonkim888.github.io

education

Korea Advanced Institute of Science and Technology (KAIST) Mar 2023 – Present

PhD Candidate in Mechanical Engineering Daejeon, Korea

Research: Design for Manufacturing, Design Optimization, Generative Design
Advisor: Prof. Namwoo Kang, Smart Design Lab
Teaching Assistant: ME203 Mechatronics System Design, ME231 Mechanics of Materials, HSS190 Freshman Seminar

University of Bath Sep 2020 – Aug 2022

M.Sc. in Data Science United Kingdom

Thesis: Fine-scale Synthetic Shape Generation with Upsampling Network
Advisor: Dr. Wenbin Li

University of Bristol Sep 2017 – Jun 2020

B.Eng. in Mechanical Engineering United Kingdom

Thesis: Testing the Limits of Generative Design for Road Bicycle Frames

experience

Narnia Labs Jan 2025 – Present

AI Researcher Seoul, Korea

Leading agentic AI development toward more autonomous engineering workflows
Originated the concept for AslanX Flow and built its first MVP, then helped shape its subsequent product pivot
Translating mechanical engineering expertise and engineering data into AI workflows for model development, evaluation, and deployment

KAIST Smart Design Lab Sep 2022 – Feb 2023

Research Intern Daejeon, Korea

Conducted preliminary research on deep generative models for engineering design optimization

SLB (previously Schlumberger) Jun 2021 – Aug 2022

Data Scientist Intern United Kingdom

Developed AI models for drill bit design optimization and recommendation
Optimized predictive steering parameters using data analysis and machine learning
Created predictive maintenance systems for drilling electronics

publications

* denotes equal contribution

Journal

Deep Generative Model-based Synthesis Framework of Four-bar Linkage Mechanisms with Target Conditions Lee, S.*, Kim, J.*, Kang, N. Journal of Computational Design and Engineering, 11(5), 318–332, 2024

Preprints

Deep Generative Design for Mass Production Kim, J.*, Kwon, Y.*, Kang, N. arXiv:2403.12098, 2024

Performance Comparison of Design Optimization and Deep Learning-based Inverse Design Jwa, M.*, Kim, J.*, Shin, S., et al. arXiv preprint, 2023

Conferences

A Study on 3D Topology Optimization Shape Reconstruction with CSG-Based Deep Learning Park, J., Kim, J., et al. KSME Annual Conference (Poster), 2025

Deep Generative Design for Manufacturing: Meeting Design Constraints of Casting and Injection Molding Kim, J., Kwon, Y., Kang, N. Asian Congress of Structural and Multidisciplinary Optimization (ACSMO), 2024

2D Diffusion Model-based 3D Design Optimization for Mass Production Kim, J., Kwon, Y., Kang, N. KSME Annual Conference, 2024

Deep Generative Design for Mass Production Kim, J., Kwon, Y., Kang, N. KSME CAE Division Spring Conference, 2024

Deep Generative Model-based Synthesis of Four-bar Linkage Mechanisms Considering Kinematic and Dynamic Conditions Lee, S.*, Kim, J., Kang, N. ASME IDETC, 2023

Deep Learning-based Parametric Inverse Design Considering Engineering Performance and Additive Manufacturing Kim, J., Lee, S., Kang, N. KSME Annual Conference, 2023

skills

Programming Python, JavaScript/TypeScript, MATLAB

ML/DL PyTorch, TensorFlow/Keras, scikit-learn

CAD/Simulation Autodesk Fusion 360, ANSYS Suite, STAR-CCM+

DevOps Git, Docker, AWS/GCP/Azure, Linux

Web React, Vite, REST API

Languages Korean (Native), English (Fluent), Spanish (Proficient)