JUIWEI FU - FU JUI-WEI | Professional Portfolio

Introduction

My name is FU, JUI-WEI, currently pursuing a master’s degree at the Graduate Institute of Networking and Multimedia, National Taiwan University (NTU).

In terms of practical experience, I specialize in full-stack web development. I led the development of several student community platforms, including NTUST GPA SEARCH and MyNTUST, both of which became the most widely used tools among NTUST students, accumulating over 124,000 annual views. My technical expertise also enabled me to participate in an industry-academia collaboration project with the Industrial Technology Research Institute (ITRI) in my junior year.

My undergraduate thesis, conducted in collaboration with ITRI, focused on “Abnormal Prediction of 5G Base Stations”. By utilizing 3GPP-defined performance indicators as input features for an LSTM model, we achieved an accuracy rate of 93% and an ROC-AUC score of 0.868. The final research results were submitted to ITRI in September 2023.

My current graduate research explores the capabilities and limitations of Large Language Models (LLMs) in both Single-Agent Systems (SAS) and Multi-Agent Systems (MAS). I am investigating how LLMs reason, collaborate, and reach consensus across multiple agents within MAS architectures. This research is geared toward applying such systems to tasks like problem solving and world simulation, with the aim of demonstrating how MAS frameworks can outperform single-agent setups in reasoning accuracy and adaptability.

Education

M.S., Graduate Institute of Networking and Multimedia

National Taiwan University, Taiwan (2024/01 - Present)
B.S., Computer Science and Information Engineering

National Taiwan University of Science and Technology, Taiwan (2020/09 - 2024/01)

Work Experience

Technician

Far Eastern Memorial Hospital · 13 March 2023 - 16 March 2025
- Emergency medical technician in ER (Emergency Room).
- Disaster medical assistance team member.
- Basic life support provider.
Software Engineer (Intern)

You-shang Technical Corp. · 01 February 2022 - 28 February 2023
- Developed motion detection software for golf swings.
- Integrated high-frame-rate camera support.
- Enhanced golfer performance analytics.