I am an AI Engineer solving real-world problems with state-of-the-art LLMs, deep learning methods, and data analysis. During my time in academia, my main research area was Information Retrieval and RAG systems under diverse language scenarios β such as Cross- and Mixed-Language Information Retrieval. Drawing on these diverse research experiences, I now focus on building and serving AI services that deliver meaningful value to the people who use them.
π CV (Updated on 2026.09.10) Β· LinkedIn Β· Google Scholar Β· ACL Anthology
-
AI Engineer, DT Innovation Team, LINA ONE (2026.01 - Present)
- Serving and optimizing LLM, TTS, and STT models for AI TMR, an AI voice agent that carries out insurance telemarketing consultations, and continuously improving them while handling issues raised from the live customer service.
- Participating in AI PS, which designs insurance plans for human TMR, analyzing insurance contract data and developing target premium calculation logic to grow APE (Annualized Premium Equivalent).
-
Research Intern, SK Telecom (2025.10 - 2025.12)
- Joined the 1st phase of the Sovereign AI Foundation Model project (a.k.a. λ νλͺ¨) and took charge of curating training data in SFT and DPO formats for A.X K1 β synthesizing data and filtering it for quality β targeted at benchmarks such as Tau-bench and KoBALT, which contributed to passing the 1st phase evaluation. Also built vibe-coding task training data with trajectories to improve A.X K1.
-
Internship, KETI (2022.04 - 2022.11)
- Participated in a national R&D project and built a Korean sentiment analysis module, covering the full cycle from collecting and curating the sentiment data to post-training and evaluating a KE-T5βbased model.
- Contributed to the projectβs quantitative outcomes with a domestic conference presentation (HCLT 2022) and a patent application.
-
M.S. in Graduate School of Artificial Intelligence, POSTECH (2023.02 - 2025.02)
- Lab: POSTECH NLP Lab
- Advisor: Gary Geunbae Lee
- Co-Advisor: Jungseul Ok, Yunsu Kim
- Master Thesis: Leveraging Code-Switched Data to Improve Multi-Vector Retrievers for Cross-Language Information Retrieval
-
B.S. in Information Convergence, Kwangwoon University (2017.03 - 2023.02)
- Major : Data Science
- Keywords : Information Retrieval, RAG, Re-ranking, Voice-Agent System, Agentic System, Cross-lingual Transfer, Code-Switching
- SK Telecom A.X Team (incl. Jonghwi Kim). (2026). A.X K1 Technical Report. (arXiv preprint). [Paper]
- Jonghwi Kim, Deokhyung Kang, Seonjeong Hwang, Yunsu Kim, Jungseul Ok, Gary Geunbae Lee. (2025). MiLQ: Benchmarking IR Models for Bilingual Web Search with Mixed Language Queries. (EMNLP 2025 main). [Paper]
- Daehee Kim, Deokhyung Kang, Jonghwi Kim, Sangwon Ryu, Gary Geunbae Lee. (2025). GuRE: Generative Query REwriter for Legal Passage Retrieval. (NLLP Workshop 2025). [Paper]
- San Kim, Jonghwi Kim, Yejin Jeon, Gary Geunbae Lee. (2025). Safeguarding RAG Pipelines with GMTP: A Gradient-based Masked Token Probability Method for Poisoned Document Detection. In Findings of the 62nd Annual Meeting of the Association for Computational Linguistics (ACL 2025 findings). [Paper]
- Heejin Do*, Sangwon Ryu*, Jonghwi Kim, Gary Geunbae Lee. (2025). Multi-Facet Blending for Faceted Query-by-Example Retrieval. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (ACL 2025 main). [Paper]
- Jonghwi Kim, Yunsu Kim, Gary Geunbae Lee. (2023). ColBERT with Adversarial Language Adaptation for Multilingual Information Retrieval. In Proceedings of the Annual Conference on Human and Cognitive Language Technology (HCLT 2023 oral), pp. 239β244. [Paper]
- Jonghwi Kim, Saim Shin, Jin Yea Jang. (2022). A Clustering-based Undersampling Method to Prevent Information Loss from Text Data. In Proceedings of the Annual Conference on Human and Cognitive Language Technology (HCLT 2022 oral), pp. 251β256. [Paper]
- Reviewer: MRL Workshop