Korea Institute of Science and Technology (KIST) / UST
Researcher (Engineering Focus) · 2021 – Present · Seoul, Korea
You can contact me on e-mail.
alcyonesha@gmail.com · Seoul, South Korea
Summary
AI engineer experienced in LLM agent systems, large-scale simulation, and cross-organizational technical advisory. Research background enables rapid adoption of state-of-the-art methods into engineering workflows.
Skills
Languages: Python, JavaScript, TypeScript
AI / ML: LLM Agents, RAG, Prompt Engineering, Fine-tuning, Multi-Agent Simulation, Evaluation Frameworks, Causal Inference
Frameworks: LangChain
Tools: Git, Docker
Communication: Korean (Native), English (Professional)
Experience
COVID-19 Policy Simulation
Commissioned by KDCA
Mapped nationwide mobility from telecom and credit card data across 50M individuals; ran large-scale agent-based simulation to model transmission dynamics, directly informing the government's 5-person gathering restriction policy.
National HIV Prevalence Modeling
MOU with KDCA
Accessed sensitive patient records under MOU with KDCA; constructed a network model of infection pathways and built an ensemble AI model to estimate national HIV prevalence; paper currently under minor revision.
Federated Computing Platform
KRW 4B government R&D
Architected a federated learning platform for privacy-preserving analysis of citizen health data collected from Seoul city-operated wearable devices across distributed data silos.
AI Alignment & Safety Research
KRW 9.9B government R&D
Serving as technical advisor to an enterprise HR team on LLM integration for personnel assessment; developing causal modeling approaches to improve interpretability of AI-assisted organizational decisions.
Multi-Party Conversation Agent (MPCA)
Extracted behavioral parameters from 8,900+ IRC conversations; developed LLM evaluation pipeline via LLM APIs to measure agent naturalness against human communication baselines. — GitHub
Persona-Driven Agent Simulation
Built multi-agent simulation via LLM APIs where personas navigate 25 workplace scenarios through agentic loop with tool use, RAG, and NPC graph interaction; 9-dimension behavioral scoring with Cohen's Kappa and Krippendorff's Alpha. HCI collaboration with Cornell HCI Lab. MOU signed with POSCO HR for deployment in new-hire assessment workflow.
Education
University of Science and Technology (UST) / KIST
Ph.D. Candidate, AI & Robotics · 2024 – Present · GPA: 4.0 / 4.5
University of Science and Technology (UST) / KIST
M.S., AI & Robotics · 2022 – 2024 · GPA: 4.0 / 4.5
Seoul National University of Science and Technology
B.S., Chemical and Biological Engineering · 2018
Publications & Presentations
"Integrated Disease Model Considering Mutation-Induced Infection Waves with COVID-19 Cases" — PLOS ONE, 2026 (first author)
"How to Incorporate Mutation-Induced Infection Waves of COVID-19?" — Society for Mathematical Biology, 2023
"Multi-Agent Combat Analysis Simulation Considering Terrain and Squad Operations" — IAMSEC, 2024
1 paper under revision
Guest Lecture — Hankyong National University, Oct 2025: Delivered lectures on LLM architectures and generative AI application patterns to undergraduate students.