Hello! I am Yuxiang Liu, a master’s student in Software Engineering at Beijing Jiaotong University, advised by Haiming Liu.

I received my Bachelor of Engineering degree from Nanjing University of Science and Technology, where I was advised by Associate Professor Cangqi Zhou.

My research interests primarily focus on large language models (LLMs) and natural language processing (NLP).

Recently, I have been following research advances in agents and harness engineering, and I am seeking agent-related internship opportunities.

Beijing Jiaotong University

Software Engineering · Master’s program

2026.09 – 2028.06

Nanjing University of Science and Technology

Computer Science and Technology · Bachelor’s degree

2022.09 – 2026.06
Average score: 82.55 · GPA: 3.22/4 · CET-4: 583 · CET-6: 508 · IELTS: 7.0

Intern
Nanjing Desay SV Automotive Electronics Co., Ltd.
Embedded Software Testing Intern · Jul 2024 – Aug 2024
NJ_SIT (Intelligent Applications R&D Department)
  • Performed embedded software testing for projects including Xiaomi and BYD, completing unit tests for over 2,500 functions with 100% coverage.
  • Used TESSY and the NXP embedded development IDE for unit testing, developing familiarity with embedded testing workflows and standards.

Honors

13th MathorCup University Mathematical Modeling Challenge (2023)First Prize
Mathematical Contest in Modeling (MCM), 2024Honorable Mention
6th National Collegiate Curling AI Challenge (2024)Third Prize
15th CP Cup National Collegiate Market Research and Analysis Competition (2025)First Prize, Jiangsu Province
China Undergraduate Mathematical Contest in Modeling (CUMCM)(2024)Second Prize, Jiangsu Province
14th MathorCup Mathematical Applications Challenge (2024)Second Prize, Jiangsu Province

NJUST Outstanding Student Scholarship · Second Class
Awarded once · 24-25Q1
NJUST Outstanding Student Scholarship · Third Class
Awarded three times · 22-23Q1, 23-24Q2, 24-25Q2

LinkRadar key node dashboard screenshot
LinkRadar

A network analysis project combining exact Brandes betweenness centrality and parallel multiprocessing with node rankings, structural failure experiments, and interactive visualizations to explore shortest-path brokerage in real networks.

Live demo: https://lyx0902.github.io/LinkRadar/

AchievementAnalysis interface screenshot
AchievementAnalysis

Overall project: Integrates Steam data from multiple sources to analyze game trends, clustering, reviews, and player engagement, providing explainable recommendations and interactive presentations.

My contribution: Implemented automated achievement classification and mainline progression analysis, interpreted player engagement using gameplay, playtime, and reviews, and visualized the data.

Live demo: https://www.soyorinlove.cn/#research/retention

ApexLogic interface screenshot
ApexLogic

A multi-agent deep research system combining multi-source retrieval and iterative review, with checkpoint-based research resumption, cross-task memory reuse, and concurrent background execution.

2023.10
7th Jiangsu Provincial College Student Arts Exhibition
First Prize, Instrumental Ensemble Division
2022.10
NJUST Freshman Basketball Tournament
Third Place