CV
Education, experience, awards, and skills. The PDF version is linked above.
Contact Information
| Name | Jihwan Kim |
| Professional Title | Integrated M.S./Ph.D. Student, KAIST |
| jihvvan.kim@kaist.ac.kr |
Professional Summary
Applied AI researcher building LLM agents and the evaluation frameworks that keep them honest. Interested in agentic task execution, multimodal and speech language models, and enterprise-grade domain-specific agents.
Experience
-
2025 - 2026 Seongnam, South Korea
WBL Residency Intern
NAVER Cloud
- Developed a domain-specific cross-lingual RLVR data synthesis method to mitigate the multilingual performance gap.
- Developed a comprehensive evaluation pipeline for Audio LLM performance, achieving high-fidelity assessment across both audio understanding and generation tasks.
- Implemented multimodal benchmarking methodologies for HyperCLOVA X Omni, enhancing the technical reportβs depth by validating model capabilities against industry-standard metrics.
- Conducted analysis on current Audio LLM research literature, identifying key architectural trends and optimization opportunities for next-generation multimodal agents.
- Contributed to the HyperCLOVA X technical report as part of a WBL project.
Education
-
2024 - Present South Korea
Integrated M.S./Ph.D.
Korea Advanced Institute of Science and Technology (KAIST)
Artificial Intelligence
-
2018 - 2024 Suwon, South Korea
Awards
-
2022 Seah Haiam Scholarship
Seah Haiam Foundation
Selected as a recipient for three consecutive years (2022β2024) based on academic merit and future potential, recognizing a consistent track record of excellence in technology and innovation.
-
2022 Dean's List (Spring 2022)
College of Software Convergence, SungKyunKwan University
Awarded for outstanding academic excellence, ranking within the top 4% of the undergraduate student body in the College of Software Convergence.
Skills
Languages & Frameworks: Python, PyTorch, Transformers
Research: LLM Agents, Benchmark Design, Multimodal Evaluation, Speech Language Models, Retrieval, RLVR
Languages
Korean : Native speaker
English : Proficient