Experience
Research & Professional Experience
AI Engineer, Causal Inference Team · NAVER · Aug 2025 – Present
Seongnam, Republic of Korea
- Research on off-policy evaluation under deterministic online ad auctions.
- Exploring kernel-based relaxation methods to reduce variance in importance sampling for deterministic eCPM-based ranking policies.
- Designing large-scale training and evaluation pipelines leveraging NAVER’s ad logs for more reliable offline policy evaluation of production models.
Machine Learning Engineer, AI Team · Dable · Mar 2022 – May 2025
Seoul, Republic of Korea
(Technical research personnel — alternative military service)
- Led a team to develop machine learning models for conversion rate (CVR) prediction, addressing data bias and scarcity in e-commerce settings.
Resolving Data Bias with Causal Inference
- Developed a novel causal inference-based methodology for CVR prediction, imputing conversion labels for non-clicked samples via counterfactual inference.
- Achieved a 2% improvement in offline CVR estimation and a 17% boost in online CVR over state-of-the-art algorithms.
- Published at IEEE ICDM 2025 (Best Paper Award Finalist).
Addressing Data Scarcity with Multi-Task and Self-Supervised Learning
- Employed multi-task learning and self-supervised learning by augmenting embedding vectors and implementing contrastive loss functions.
- Achieved a 14% improvement in online CVR estimation under data-limited conditions.
Master’s Student, Information System Lab · KAIST · Feb 2019 – Feb 2022
Daejeon, Republic of Korea
Matrix Completion with Hierarchical Graph Side Information
- Researched solutions to the cold-start problem and data scarcity in matrix completion using social graphs.
- Theorized the information-theoretic limit for the minimum observed matrix entries needed for perfect recovery.
- Developed a computationally efficient algorithm achieving an 18% reduction in prediction error and an 80% decrease in computational time, while meeting the information-theoretic limit.
- Published at NeurIPS 2020, ISIT 2022, and IEEE Transactions on Information Theory 2024.
Medical Image-Based Patient Search System
- Collaborated with Samsung Medical Center on AI-driven medical diagnostics.
- Developed a deep learning retrieval system for chest disease prediction using PyTorch.
- Achieved 85% accuracy in classifying 14 types of chest diseases.
Research Intern, ML Infra Lab · SK Telecom · Mar 2018 – Aug 2018
Seongnam, Republic of Korea
- Designed a dynamic learning rate schedule, accelerating model convergence by 3× while preserving accuracy.
- Designed a distributed deep learning training system, reducing computational costs by 30%.
Education
University of Illinois Urbana-Champaign (UIUC)
Ph.D. in Information Science · Starting Fall 2026
Korea Advanced Institute of Science and Technology (KAIST)
M.S. in Electrical Engineering · Feb 2019 – Feb 2022
GPA: 3.91 / 4.3 · Advisor: Prof. Changho Suh
Thesis: “Leveraging Hierarchical Similarity Graphs for Matrix Completion”
Korea Advanced Institute of Science and Technology (KAIST)
B.S. in Electrical Engineering · Mar 2015 – Feb 2019
GPA: 3.93 / 4.3 · Magna Cum Laude
Honors & Awards
- Best Paper Award Finalist, IEEE ICDM 2025
- Magna Cum Laude, KAIST (Feb 2019)
- KAIST Support Scholarship (Mar 2019 – Dec 2020) — National full scholarship, M.S. program ($24,000)
- National Science & Technology Scholarship (Mar 2015 – Dec 2018) — National full scholarship, B.S. program ($15,000)
- KAIST Support Scholarship (Mar 2015 – Dec 2018) — 8 semesters, B.S. program ($20,000)
Technical Skills
Programming Languages: Python, MATLAB, JavaScript, TypeScript, C++
ML Frameworks: PyTorch, TensorFlow, Scikit-learn
Data & Visualization: Pandas, NumPy, Matplotlib, Seaborn
Big Data & Distributed: Spark, Hadoop, Ray
Database & Cloud: SQL, AWS
DevOps: Git, Docker, Kubernetes
Languages: English (TOEFL iBT 106), Korean (Native)