// intro
About Me
I build AI systems that help people communicate across languages and cultures — at the intersection of research and production.
As a Staff Engineer at Samsung Electronics, I'm part of the team building machine translation for Galaxy AI, working alongside global research labs to ship multilingual AI features to hundreds of millions of devices worldwide. My current focus is on automation — building data pipelines, quality inspection systems, and CI/CD infrastructure that make model iteration faster and more reliable.
I hold a Ph.D. in Computer Science from Korea University's NLP & AI Lab, where my research on dialogue systems — grounding conversations in persona and world knowledge — was published at EMNLP, AAAI, COLING, and ACL. Along the way I collaborated with Naver, Hyundai Motor, and NCSoft on real-world NLP challenges that shaped my belief that the best research ships.
// updates
News
2026
PaperPaper accepted at ACL 2026 — "I Don't Need Solution. I Need Emotional Support: Empathetic LLMs based on Emotional Validation" (2nd author)
09/2024
JoinedJoined Samsung Electronics MX as Staff Engineer, Galaxy AI
07/2024
ReviewInvited to serve as Reviewer for EMNLP 2024
03/2024
AwardReceived Integrated Talent Scholarship from Brain Korea 21 (BK21)
02/2024
PaperPaper on CED (Critical Error Detection) accepted at LREC-COLING 2024
10/2023
Paper3 papers accepted at EMNLP 2023
09/2023
ReviewInvited to serve as Reviewer for Expert Systems With Applications
06/2023
ReviewInvited to serve as Reviewer for EMNLP 2023
01/2023
ReviewInvited to serve as Reviewer for ACL 2023
10/2022
Hosted 1st Workshop on Customized Chat Grounding Persona and Knowledge at COLING 2022
07/2022
ReviewInvited to serve as Reviewer for EMNLP 2022
// focus areas
Research Interests
Dialogue Systems
Machine Translation
Relation Extraction
Vision & Language
Visual Reasoning
Commonsense Knowledge
Semantic Representation
Large Language Models
// work experience
Industry
SE
Staff Engineer — Galaxy AI, Machine Translation
Samsung Electronics · MX Division
2024.09 — Present
- Developing multilingual neural machine translation models powering Galaxy AI features — including Live Translate (calls), Interpreter, Writing Assist, Note Assist, and Browsing Assist — deployed across Galaxy smartphones and tablets worldwide
- Collaborating with overseas research labs to research and align multilingual MT models across diverse language pairs and regional requirements
- Building a data augmentation pipeline to automatically generate and diversify training data for low-resource language pairs
- Designing automated data inspection workflows for quality control and noise filtering at scale
- Implementing CI/CD pipelines for automated model build, evaluation, and deployment — reducing iteration cycles and enabling continuous model improvement
SF
Research Collaborator
Samsung Fire & Marine Insurance
2023.07 — 2023.12
- Explored new digital services for enhancing digital competitiveness
- Collaborated with UX/UI designers to introduce new services via MVP model
N
Research Collaborator
Naver Corporation
2022.07 — 2023.03
- Built the first Korean document-level relation extraction dataset from Naver's encyclopedia
- Developed coreference-resolution-enhanced models for entity relation classification
H
Research Collaborator
Hyundai Motor Company
2021.06 — 2022.03
- Developed intent classification and slot-filling model for in-vehicle dialogue
- Applied multi-task training on language models for simultaneous intent and slot prediction
NC
Research Collaborator
NCSoft Company
2020.06 — 2022.12
- Created FoCus — a dialogue dataset grounding user persona and knowledge via Amazon MTurk
- Built dialogue system informing users about Wikipedia knowledge based on persona
// selected works
Publications
International Conference
ACL
I Don't Need Solution. I Need Emotional Support: Empathetic LLMs based on Emotional Validation To Appear
The 63rd Annual Meeting of the Association for Computational Linguistics (ACL 2026)
LREC-COLING
KNOTICED: A Dataset for Critical Error Detection in English-Korean Machine Translation
LREC-COLING 2024
EMNLP
EMNLP 2023 Findings
EMNLP
EMNLP
EMNLP
COLING
COLING 2022
AAAI
Towards Filling the Gap between Written and Spoken Dialogues for Multi-Domain Dialogue State Tracking
AAAI 2022 — DSTC10 Workshop
AAAI
EMNLP
EMNLP 2021 — NewSum Workshop
COLING
COLING 2020
International Journal
Journal
An Analysis on Language Transfer of Pre-trained Language Model with Cross-lingual Post-training
Expert Systems with Applications (IF: 7.5) · 2025
Journal
Applied Intelligence (IF: 5.3)
Journal
IEEE/ACM Transactions on Audio, Speech and Language Processing (IF: 4.364)
Journal
Applied Sciences, 2022, 12(23)
Journal
Applied Sciences, 2022, 12(19)
Journal
Applied Sciences, 2022, 12(18)
Journal
Applied Sciences, 2021, 11(23)
Domestic Conference & Journal
HCLT
Enhancing Empathetic Response Generation in Large Language Models Based on Emotional Validation Theory
Annual Conference on Human and Language Technology (HCLT 2024)
HCLT
Audio-Visual Scene Aware Dialogue System Utilizing Action From Vision and Language Features
HCLT 2023
HCLT
Dialogue Relation Extraction using Dialogue Graph Best Paper
HCLT 2022
HCLT
Analysis on Question Understanding of Language Models using Clever Hans Tests
HCLT 2021
HCLT
Commonsense Graph Path Learning Model for OpenBook Question and Answering Best Paper
HCLT 2020
HCLT
Hyperpartisan news Detection using Sentence BERT Embedding
HCLT 2019
Journal
Journal of the Korea Convergence Society
// research
Research Projects
MIRI.DIH — Image Retrieval Service
2024.02 — 2024.08
- Developed image retrieval services for end users
- Explored relative image components for canvas-based interaction
Korean Document-level Relation Extraction · Naver
2022.07 — 2023.03
- Built first Korean document-level RE dataset from Naver Encyclopedia documents
- Model combining coreference resolution, entity typing, and relation extraction
In-Vehicle Dialogue System · Hyundai Motor
2021.06 — 2022.03
- Intent classification and slot filling for Hyundai vehicle dialogue
- Multi-task training for simultaneous intent and slot prediction
FoCus Dialogue Dataset · NCSoft
2020.06 — 2022.12
- Created FoCus dataset grounding user persona and knowledge via Amazon MTurk
- Built dialogue system informing users about Wikipedia knowledge of landmarks
// recognition
Awards & Honors
Best Research Award — Korea University
2023
Best Paper Award — 34th Annual Conference on Human and Language Technology (HCLT)
2022
3rd Place — DSTC10 Track 2, Knowledge-grounded Task-oriented Dialogue on Spoken Conversations
2021
Best Paper Award — 32nd Annual Conference on Human and Language Technology (HCLT)
2020
5th Place — 8th Financial Style Visualization Competition (Korea Credit Bureau)
2019
Excellence Award — 5th L.POINT Big Data Competition
2019
Encouragement Award — 4th Meteorological Agency Big Data Contest
2018
// background
Education
Korea University
M.S & Ph.D in Computer Science and Engineering
NLP & AI Lab · Advisor: Heuiseok Lim
2019.09 — 2024.08
Sungkyunkwan University
B.S in Library and Information Science & Data Science
2015.03 — 2019.08
// reach out