Hulin Wang

Hulin Wang

Ph.D. Candidate · LLM Agents for Vulnerability Discovery & Repair

SEFCOM Security Lab @ Arizona State University

I’m a Ph.D. candidate in the SEFCOM Security Lab at Arizona State University, advised by Dr. Tiffany Bao, Dr. Fish Wang, Dr. Adam Doupé, and Dr. Yan Shoshitaishvili. I build LLM agents that find, triage, and fix real-world software vulnerabilities.

I led the patching component of Team Shellphish’s autonomous cyber reasoning system, which placed 5th worldwide in DARPA’s AI Cyber Challenge (AIxCC). In summer 2026, I was an Applied Scientist Intern at Amazon, building agentic AI systems to tackle security challenges at scale.

I’m looking for full-time Security Researcher and Research Engineer roles, starting after I graduate in Fall 2027.

Download my résumé.

Interests
  • LLM Agents for Security
  • Automated Vulnerability Repair
  • Fuzzing & Program Analysis
  • Firmware & Cyber-Physical Security
Education
  • Ph.D. in Computer Science, Expected Fall 2027

    Arizona State University

  • M.S. in Computer Science, 2022

    University of Southern California

  • B.S. in Computer Science, 2020

    University of California, Davis

Experience

 
 
 
 
 
Applied Scientist Intern
Amazon
May 2026 – Aug 2026 New York, NY
Built multi-agent LLM systems to triage static-analysis findings for security rule development.
 
 
 
 
 
Security Researcher & Ph.D. Candidate
Aug 2022 – Present Tempe, AZ
Research on LLM agents for vulnerability repair, LLM-assisted fuzzing, and UAV firmware security (DARPA FIRE).
 
 
 
 
 
Patching Component Lead
Aug 2023 – Aug 2025 Tempe, AZ
Led the LLM-powered patching component of Shellphish’s autonomous cyber reasoning system — 5th place in DARPA AIxCC.
 
 
 
 
 
Student Researcher, IT Governance, Risk & Compliance
University of Southern California
Oct 2021 – May 2022 Los Angeles, CA
Built risk metrics and monitoring for university-wide IT infrastructure.
 
 
 
 
 
Software Engineer Intern
LitePoint
Jun 2019 – Sep 2019 San Jose, CA
Built a Python ETL pipeline for manufacturing data and optimized SQL Server queries.

Honors

5th Place Worldwide, DARPA AI Cyber Challenge (AIxCC)
Patching Component Lead. Also a 2024 finalist ($2M award).
DEF CON CTF Finals
Competed in the finals with Team Shellphish in 2023 and 2025.
SCAI Doctoral Fellowship & Fulton Fellowship
PLMW@PLDI 2022 Scholarship

Publications

Under review
Root-Cause-Driven Automated Vulnerability Repair. First author
SeedSmith: LLM-Driven Seed Synthesis for Directed Fuzzing.
(2026). AIJon: Automated Generation of Annotations for Fuzzing. arXiv.

PDF

(2026). BuildBench: Benchmarking LLM Agents on Compiling Real-World Open-Source Software. TMLR 2026.

PDF

(2022). FedNLP: Benchmarking Federated Learning Methods for Natural Language Processing Tasks. NAACL Findings 2022.

PDF Cite DOI

Service

  • Artifact Evaluation Committee: IEEE S&P 2026; ACM CCS 2024 (Distinguished Artifact Reviewer)
  • Journal Reviewer: IEEE Transactions on Dependable and Secure Computing (TDSC); ACM Transactions on Intelligent Systems and Technology (TIST, external reviewer)