Shaolin Xie

Shaolin (Shelly) Xie

PhD Student

University of Southern California

Biography

I am a PhD student in computer science at the University of Southern California, advised by Prof. Ibrahim Sabek in the NEXDIG Lab. My research focuses on building next‑generation data systems and multi‑agent systems. I am especially interested in lifelong query optimization, geospatial reasoning, and collaborative data management.

I am always open to collaborations and discussions. Feel free to reach out if you are interested in data systems, multi‑agent systems, or related topics.

Interests

  • Multi‑agent systems
  • Query optimization and database performance
  • Geospatial data systems

Education

  • Ph.D. in Computer Science (in progress)
    University of Southern California
  • M.S. in Computer Science
    Boston University
  • B.S. in Computer Science and Data Science
    Worcester Polytechnic Institute

Recent News

All news »

[Nov. 2025] Presented Evaluating Intrinsic Geospatial Topological Reasoning in LLMs at the GeoGenAgent’25 workshop.

[Jun. 2025] Our paper LIMAO: A Framework for Lifelong Modular Learned Query Optimization is accepted to VLDB'25.

Recent Publications

Evaluating Intrinsic Geospatial Topological Reasoning in LLMs [SIGSPATIAL - GeoGenAgent’25]

Shaolin Xie, Shang‑Ling Hsu, Qihan Zhang, Yiming Gao, Cyrus Shahabi, Ibrahim Sabek

Paper
LIMAO: A Framework for Lifelong Modular Learned Query Optimization [VLDB 2025]

Qihan Zhang, Shaolin Xie, Ibrahim Sabek

Paper
QuIT your B+ Tree for Quick Insertion Tree [EDBT 2025]

Aneesh Raman, Konstantinos Karatsenidis, Shaolin Xie, Matthaios Olma, Subhadeep Sarkar, Manos Athanassoulis

Paper

Participate

How good is this analysis?

Three agents explored the same datasets, each leaving behind chains of findings, and an LLM judge scored every chain on six criteria. This study asks you the same six questions about a random sample of them — with exactly the content the judge was shown — so that human and model judgements can be compared directly.

About three minutes per trajectory — the survey says how many on its first screen. No sign‑up; nothing is sent anywhere until you choose to download and send your answers.

Take part

Contact

© 2025 Shaolin Xie