Customer-Facing AI

Interactive AI

Project 1 A Design of Human-AI Interaction based Technological Predicting Process

This project explores how large language models (LLMs) can support human-centered patent analysis and technology forecasting. Through experiments in trend analysis, affiliate mapping, and topic labeling, the team investigates the practical utility of LLMs in extracting insights from patent data. The project also focuses on developing intuitive interfaces and topological mapping techniques that align patents with relevant products or services, enabling more effective detection of innovation gaps. (PI: Myunghwan Yun, SNU)

Publications & Presentations

• Park, S., Kim, G., Lee, S. (2024) Evaluating the Utility of LLMs in Patent Analyses: Focusing on Technology Trend Analyses, Affiliate Integration, Patent Search Query Formulation, and Topic Labeling.

---

Project 2 Liquid Metal-based Mechanical/Electric Signal Measuring Sensor Development

This project aims to develop flexible, self-powered sensors by integrating liquid metal and thermoelectric materials, enabling real-time health monitoring through wearable devices. The team designed stretchable thermoelectric generators and laser-sintered conductors to collect bio-signals such as blood pressure without traditional cuffs. Combined with AI-driven data processing and wireless communication, the system paves the way for robust and energy-efficient healthcare wearables. (PI: Seung Hwan Ko, SNU)

Publications & Presentations

• Zadan, M., Wertz, A., Shah, D., Patel, D., Zu, W., Han, Y., Gelorme, J., Mea, H.J., Yao, L., Malakooti, M., Ko, S.H., Kazem, N., Majidi, C. (2024) Stretchable Thermoelectric Generators for Self-Powered Wearable Health Monitoring (Learn more)
• Park, J.J., Hong, S., Jung, Y., Won, P., Majidi, C., Kim, M., Ko, S.H. (2024) Highly Sensitive Cuffless Blood Pressure Monitoring with Selective Laser-sintered Liquid Metal Conductors. (Learn more)

---

Project 3 Customized Assistance Technology Development with Individual Situations

This project develops optimization algorithms based on LLMs to support complex, context-sensitive decision-making. By modeling user intent through parameterization and dialogue, the system adapts to various domains such as travel planning or healthcare support. The research aims to establish a scalable, general-purpose AI decision support system capable of operating under ambiguity and personalization constraints. (PI: Joonhwan Lee, SNU)

Publications & Presentations

• Lee, K., Eun, J., Kang, S., Jeon, S., Song, C., Lee, J. (2025) LLM-Powered Decision Support: A Parameter Optimization Approach

Participating Researchers