AI-built test environments
Developing agents that read service connector code and documentation to generate resettable test environments. The workflow retains the existing SOAR platform and connector while using a generated external service.
CURRICULUM VITAE
Pei-Yu Tseng.
Ph.D. candidate in Informatics
The Pennsylvania State University · State College, PA
jerry950909@gmail.com · paiyutsutw.github.io/my_website
Research interests
Ph.D. candidate in Informatics at Penn State, advised by Professor Peng Liu. Research in AI agents for security automation, with a background in threat intelligence and security engineering.
LLM agents · Security automation · Benchmark design · Small-model training · Threat intelligence
Current research
Small models for security playbooks
Constructing source-grounded scenarios and behavioral scoring for playbook generation from natural-language goals and retrievable documentation. Benchmark development is underway; small-model training, including reinforcement learning, is the next research objective.
Experience
Penn State University
2023 — PresentResearch Assistant · State College, PA
Research on LLM agents for threat intelligence and security automation, including generated test environments and benchmarks for playbook generation. Advised by Professor Peng Liu.
Academia Sinica
2021 — 2023Research Assistant · Taipei, Taiwan
Graph-based APT detection, threat knowledge bases, kernel audit logs, and representation learning for honeypot traffic.
Chiayi County Government
2020Penetration Tester · Taiwan
Security assessments of government web applications and IoT infrastructure, with remediation guidance.
National Chung Cheng University
2019 — 2020Network Administrator · Taiwan
Network infrastructure, firewalls, VPN, DNS, and web hosting for the College of Engineering.
Education
The Pennsylvania State University
2023 — PresentPh.D. in Informatics · In progress
National Chung Cheng University
2019 — 2021M.S. in Computer Science and Information Engineering
National Taichung University of Education
2015 — 2019B.S. in Computer Science
Technical focus
AI & evaluation
LLM agents, retrieval-augmented generation, benchmark design, supervised fine-tuning, reinforcement learning, multi-agent systems, graph learning
Security
SOAR, SIEM, Splunk, ELK, threat intelligence, threat hunting, MITRE ATT&CK, detection engineering
Engineering
Python, PyTorch, TensorFlow, Bash, regular expressions, Linux, Windows, Docker, graph databases
Selected publications
2026
2026
2026
2025
2024
2023