Postdoctoral Researcher · KAIST

Jiho Kim

I build realistic, controllable, and scalable simulators to evaluate and train AI agents. I also design agents that reason over structured data such as knowledge graphs and electronic health records.

Postdoc at KAIST, where I completed my PhD under Edward Choi. I also work closely with Yohan Jo at Seoul National University. Previously an Applied Scientist Intern at Amazon.

Research

My research follows two complementary threads. First, I build realistic, controllable, and scalable simulators to systematically evaluate and train AI agents — for example, long-term multi-party dialogue understanding in DialSim and proactive, personalized assistants in ProPerSim. Second, I design agents that reason over structured data such as knowledge graphs and clinical records, from reliable knowledge-graph reasoning (FactKG, KG-GPT, R2-KG) to consistency verification over electronic health records (EHRCon, EHR-ReasonCon). Across both, I aim for agents that reason robustly and stay reliable in long-horizon, high-stakes settings.

01

LLM Agents

Simulators and agentic systems that reason, adapt, and interact over long-horizon, high-stakes scenarios.

02

Knowledge Graphs

Reliable, reasoning-intensive question answering and fact verification grounded in structured knowledge.

03

Healthcare

Clinical agents and patient simulators that operate safely over electronic health records and real interactions.

04

Consistency Check

Verifying agreement between unstructured notes and structured data — error detection in high-stakes records.

News

2026.02

Completed my PhD at KAIST and started a postdoctoral position.

2026.01

Wrapped up my Applied Scientist Internship at Amazon.

2026

ProPerSim accepted to ICLR 2026; DialSim to appear at the ICLR 2026 MemAgents workshop.

2025.10

Joined Amazon as an Applied Scientist Intern.

2025.09

PatientSim selected as a Spotlight at NeurIPS 2025 Datasets & Benchmarks.

Publications

* equal contribution

Towards Error-Free EHRs: Reasoning-Intensive Consistency Verification Between Clinical Notes and Structured Tables in Electronic Health Records

Yeonsu Kwon*, Jiho Kim*, Junseong Choi, Paloma Rabaey, Minseo Kim, Sujeong Im, Jeewon Yang, Jun-Min Lee, Sangji Lee, Jiwon Kim, Hangyul Yoon, Hyunwook Kwon, Edward Choi

ICML 2026 SD4H WorkshopHealthcare·arXiv ↗

DialSim

DialSim: A Dialogue Simulator for Evaluating Long-Term Multi-Party Dialogue Understanding of Conversational Agents

Jiho Kim, Woosog Chay, Hyeonji Hwang, Daeun Kyung, Hyunseung Chung, Eunbyeol Cho, Yeonsu Kwon, Yohan Jo, Edward Choi

ICLR 2026 MemAgents WorkshopLLM Agents·arXiv ↗

ProPerSim

ProPerSim: Developing Proactive and Personalized AI Assistants through User-Assistant Simulation

Jiho Kim, Junseong Choi, Woosog Chay, Daeun Kyung, Yeonsu Kwon, Yohan Jo, Edward Choi

ICLR 2026Financial Aid AwardLLM Agents·arXiv ↗

ECG-Agent

ECG-Agent: On-Device Tool-Calling Agent for ECG Multi-Turn-Dialogue

Hyunseung Chung, Jungwoo Oh, Daeun Kyung, Jiho Kim, Yeonsu Kwon, Min-Gyu Kim, Edward Choi

ICASSP 2026Healthcare·arXiv ↗

R2-KG

R2-KG: General-Purpose Dual-Agent Framework for Reliable Reasoning on Knowledge Graphs

Sumin Jo*, Junseong Choi*, Jiho Kim, Edward Choi

AACL 2025 FindingsKnowledge Graph·arXiv ↗

PatientSim: A Persona-Driven Simulator for Realistic Doctor-Patient Interactions

Daeun Kyung, Hyunseung Chung, Seongsu Bae, Jiho Kim, Jae Ho Sohn, Taerim Kim, Soo Kyung Kim*, Edward Choi*

NeurIPS 2025 D&BSpotlightHealthcare·arXiv ↗

EHRCon

EHRCon: Dataset for Checking Consistency between Unstructured Notes and Structured Tables in Electronic Health Records

Yeonsu Kwon*, Jiho Kim*, Gyubok Lee, Seongsu Bae, Daeun Kyung, Wonchul Cha, Tom Pollard, Alistair Johnson, Edward Choi

NeurIPS 2024 D&BSpotlightFinancial Aid AwardHealthcare·Project ↗

KG-GPT

KG-GPT: A General Framework for Reasoning on Knowledge Graphs Using Large Language Models

Jiho Kim, Yeonsu Kwon, Yohan Jo, Edward Choi

EMNLP 2023 FindingsKnowledge Graph·arXiv ↗

FactKG

FactKG: Fact Verification via Reasoning on Knowledge Graphs

Jiho Kim, Sungjin Park, Yeonsu Kwon, Yohan Jo, James Thorne, Edward Choi

ACL 2023Knowledge Graph·arXiv ↗

Text-To-ECG: 12-Lead Electrocardiogram Synthesis Conditioned on Clinical Text Reports

Hyunseung Chung, Jiho Kim, Joon-myoung Kwon, Ki-Hyun Jeon, Min Sung Lee, Edward Choi

ICASSP 2023Healthcare·arXiv ↗

Psoriasis

Measurement of psoriasis-affected area with artificial neural network

Woo Hyup Lee, Sungmin Lee, Jiho Kim, Ju Hee Han, Yeong Ho Kim, Junhwi Kim, Ji Hyun Lee, Chul Hwan Bang

JAAD 2023Healthcare · IF 15.49·Paper ↗

Graph-Text Multi-Modal Pre-training for Medical Representation Learning

Sungjin Park, Seongsu Bae, Jiho Kim, Tackeun Kim, Edward Choi

CHIL 2022Knowledge Graph·arXiv ↗

EHR-QA

Question Answering for Complex Electronic Health Records Database using Unified Encoder-Decoder Architecture

Seongsu Bae, Daeyoung Kim, Jiho Kim, Edward Choi

ML4H 2021Oral SpotlightHealthcare·Paper ↗

Experience

Oct 2025 – Jan 2026

AmazonApplied Scientist Intern

Education

Aug 2021 – Feb 2026

Ph.D. in AI, KAISTAdvisor: Prof. Edward Choi

Mar 2017 – Aug 2021

B.S. in Electrical Engineering & Computer Science, KAISTMinored in Computer Science

Teaching & Talks

Jul 2025

Invited Talk — EHRConKorean Society of Imaging Informatics in Medicine

Aug 2025

KoSAIM Summer SchoolHands-on: Fine-tuning a Clinical Domain LLM

Jun 2023

KAIST Lecture Series — "ChatGPT & Applications"KOOC & KAIST IT Academy

Aug 2023

KoSAIM Summer SchoolHands-on: NumPy and PyTorch Intro

Fall 2021

Teaching Assistant, KAISTProgramming for AI

Academic Service

Conference

Reviewer EMNLP 2023 · COLM 2024, 2025, 2026 · ICLR 2025, 2026 · CHIL 2025, 2026 · ICML 2026 · ACL Rolling Review (ARR): 2024 (Feb, Apr, Jun, Oct, Dec), 2025 (Feb, May, Jul, Oct), 2026 (Jan)

Journal

Reviewer ACM Health · ACM TOIS · TASLP · ESWA · AIIM · Neural Networks · Information Fusion