Shengxiang Lin

林圣翔

Zhejiang University Ph.D, Zhejiang University (2026)
Artificial Intelligence Academic Researcher
Xingqing Campus, 28 Xianning West Road, Xi'an City, Shaanxi Province

Short-bio
Hi, I am Shengxiang Lin, an incoming Ph.D. student at Zhejiang University advised by Prof. Xiaolin Zheng. I am currently completing my B.Eng. in Computer Science and Technology (Qian Xuesen Honor Program) at Xi'an Jiaotong University (2022-2026). I was a visiting student at UC Berkeley in 2024.
My research interests primarily focus on Artificial Intelligence.
Feel free to contact me if you are interested in discussing or collaborating with me.
Education
  • Zhejiang University

    Zhejiang University

    Ph.D. in Artificial Intelligence, College of Computer Science and Technology 2026.09 - 2031.06

  • University of California, Berkeley

    University of California, Berkeley

    Visiting Student, UCBX Concurrent International 2024.08 - 2024.12

  • Xi'an Jiaotong University

    Xi'an Jiaotong University

    B.S. in Computer Science and Technology (Qian Xuesen Honor Class), Faculty of Electronic and Information Engineering 2022.09 - 2026.06

Experience
  • Innovative Talents Science Training Program

    Innovative Talents Science Training Program

    Research Internship 2024.07 - 2024.09

Service
  • Member at Berkeley TechFounders’ Technology Department: 2024
  • Student Member at China Computer Federation: 2024,2025
  • Minister at XJTU Computer Science and Technology Experimental Class-Mentor-Group: 2023,2024,2025,2026
Selected Publications (view all )
All publications
Featured Projects (view all ) GitHub User's stars
Integrated Management and Intelligent Service Platform for Book Sales
Integrated Management and Intelligent Service Platform for Book Sales

2025.05 - 2025.06 Lead Developer Open Source

Implement a Douban book data scraping module to collect over 100,000+ book entries and 200,000+ user records, storing this data in a PostgreSQL relational database engineered to 3NF. The system features a Flask-RESTful server that integrates Deepseek-V3’s large model API and employs a hybrid recommendation algorithm. The functionality includes the admin panel for book CRUD operations and inventory dashboards, plus the User Portal with categorized search, shopping cart transactions, hybrid algorithm-powered recommendations, and AI-driven Q&A.

Integrated Management and Intelligent Service Platform for Book Sales
Integrated Management and Intelligent Service Platform for Book Sales

2025.05 - 2025.06 Lead Developer Open Source

Implement a Douban book data scraping module to collect over 100,000+ book entries and 200,000+ user records, storing this data in a PostgreSQL relational database engineered to 3NF. The system features a Flask-RESTful server that integrates Deepseek-V3’s large model API and employs a hybrid recommendation algorithm. The functionality includes the admin panel for book CRUD operations and inventory dashboards, plus the User Portal with categorized search, shopping cart transactions, hybrid algorithm-powered recommendations, and AI-driven Q&A.

Tencent QQ Intelligent Dialogue System
Tencent QQ Intelligent Dialogue System

2025.03 - 2025.06 Lead Developer Open Source

This project implements a distributed security service architecture, utilizing RESTful APIs for frontend-backend network interfaces and SQLx for database construction. It integrates multiple multimodal large language models such as Deepseek and Doubao, employing the LLOneBot middleware to handle request routing and dialogue state management. The system supports dynamic persona response configuration for various contexts through methods like real-time prompt modification, and has been successfully deployed and operates stably on the QQ platform.

Tencent QQ Intelligent Dialogue System
Tencent QQ Intelligent Dialogue System

2025.03 - 2025.06 Lead Developer Open Source

This project implements a distributed security service architecture, utilizing RESTful APIs for frontend-backend network interfaces and SQLx for database construction. It integrates multiple multimodal large language models such as Deepseek and Doubao, employing the LLOneBot middleware to handle request routing and dialogue state management. The system supports dynamic persona response configuration for various contexts through methods like real-time prompt modification, and has been successfully deployed and operates stably on the QQ platform.

Paranormal XJTU
Paranormal XJTU

2024.03 - 2024.04 Lead Developer Open Source

Based on XJTU's century-old history, this project transforms historical campus buildings and legends into core game narrative elements, achieving deep integration between campus culture and game content. Artistically, it employs customized pixel-art maps that fully cover the main Xingqing campus. Technically, it features a dynamically developed real-time character tracking system for automated scene transitions and a self-developed interaction engine that supports multi-character movement—including walking, running, and interactive actions—along with seamless scene rendering.

Paranormal XJTU
Paranormal XJTU

2024.03 - 2024.04 Lead Developer Open Source

Based on XJTU's century-old history, this project transforms historical campus buildings and legends into core game narrative elements, achieving deep integration between campus culture and game content. Artistically, it employs customized pixel-art maps that fully cover the main Xingqing campus. Technically, it features a dynamically developed real-time character tracking system for automated scene transitions and a self-developed interaction engine that supports multi-character movement—including walking, running, and interactive actions—along with seamless scene rendering.

All projects