Summary
I am a postdoctoral researcher at Georgia Tech, conducting research on AI-driven vulnerability discovery and patching. I received my Ph.D. in Computer Science from Korea University under the supervision of Professor Jiwon Yoon. Previously, I served as a captain in the R.O.K. Army and worked at the Agency for Defense Development (ADD) and the R.O.K. Cyber Operations Command, gaining experience in both security research and security engineering.
Work Experience
| Georgia Tech | Aug. 2025 – Present |
| Postdoctoral Researcher | Atlanta, USA |
| Korea University | June. 2025 – Aug. 2025 |
| Postdoctoral Researcher | Seoul, Korea |
| R.O.K. Cyber Operation Command | Oct. 2022 – Present |
| Security Engineer | Seoul, Korea |
- Performed vulnerability assessments for R.O.K. military IT infrastructure.
- Worked as a red team during R.O.K. military cyber operation exercises.
- Conducted cybersecurity management ability assessment for public institutions in South Korea.
- Keywords: Red team, Web hacking, Reverse engineering, Binary exploitation, APT attack
- Frameworks/Tools: Metasploit, Burp suite, IDA, WinDBG, Cobalt strike, Nmap, Python, C/C++, Powershell
| Agency For Defense Development | Jul. 2017 – Sep. 2022 |
| Security Researcher | Seoul, Korea |
Research on National-level cyberattack defense technologies
Jan. 2021 – Sep. 2022
- Goal: Organize adversaries’ cyberattack operations into attack chains, categorize them into appropriate campaigns, and respond automatically to disrupt the attacker’s ultimate goals.
- Researched predicting the next attack using Bayesian network and MITRE ATT&CK.
- Implemented network-level and host-level automatic defense using SDN.
- Keywords: APT Attack, Automatic response, MITRE ATT&CK, Bayesian Network, SDN
- Frameworks/Tools: MITRE ATT&CK, bnlearn, ONOS
- Language: Python, R, Javascript
Research on techniques for evaluating binary fuzzing results
Jan. 2018 – Oct. 2020
- Goal: Develop techniques to analyze and evaluate crashes generated from software fuzzing to identify root causes and automatically assess whether they could lead to vulnerabilities.
- Developed Linux-based taint analysis tool for Windows x64.
- Introduced crash triage technique using additional directed fuzzing and taint analysis.
- Keywords: Fuzzing, Crash triage, Crash prioritization, Root cause analysis, Dynamic binary instrumentation, Taint analysis
- Frameworks/Tools: WinAFL, libdft, WinDBG, Pintool, Dynamorio, Valgrind
- Language: Python, C/C++, Javascript
Research on cyber threat analysis and countermeasures for warship systems
Jul. 2017 – Dec. 2017
- Conducted threat analysis and proposed countermeasures for R.O.K. navy warship information systems based on NIST standards.
- Keywords: Threat analysis, Risk management, NIST SP 800-53, NIST SP 800-37, NIST SP 800-30
| Plain Bagel, Inc | Mar. 2015 – Feb. 2017 |
| Full Stack Developer(Part-time) | Seoul, Korea |
Slidee: Platform for editing and sharing YouTube video stills
Mar. 2015 – Feb. 2017
- Built a web-based editor to convert YouTube videos into screenshots with captions.
- Built a web platform to share user-generated content.
- Implemented an ELK-based user and service statistics analysis server.
- Optimized cloud hosting and databases for reliable service and cost optimization.
- Framework/Tools: React, Redux, Express.js, MongoDB, ELK stack, AWS, Google Analytics
- Language: Python, Javascript, Node.js
Education
| Korea University | Sep. 2018 – Present |
|---|---|
| Ph.D. in Information Security | Seoul, Korea |
| Advisor: Prof. Jiwon Yoon | Website |
| Korea University | Mar. 2013 – Feb. 2017 |
|---|---|
| B.S. in Cyber Defense | Seoul, Korea |
| Hansung Science High School | Mar. 2011 – Feb 2013 |
|---|---|
| Seoul, Korea |
Publications
Logs In, Patches Out: Automated Vulnerability Repair via Tree-of-Thought LLM Analysis.
Youngjoon Kim, Sunguk Shin, Hyoungshick Kim* , and Jiwon Yoon*
* Corresponding authors
USENIX Security, 2025Enhancing Graph Of Thought: Enhancing Prompts with LLM Rationales and Dynamic Temperature Control.
Sunguk Shin and Youngjoon Kim*
* Corresponding author
International Conference on Learning Representations (ICLR), 2025SCVMON: Data-oriented attack recovery for RVs based on safety-critical variable monitoring.
Sangbin Park, Youngjoon Kim, and Donghoon Lee
International Symposium on Research in Attacks, Intrusions, and Defenses (RAID), 2023BAN: Predicting APT Attack Based on Bayesian Network With MITRE ATT&CK Framework.
Youngjoon Kim, Insup Lee, Hyuk Kwon, Gyeongsik Lee, and Jiwon Yoon
IEEE Access, 2023A new approach to training more interpretable model with additional segmentation.
Sunguk Shin, Youngjoon Kim and Jiwon Yoon
Pattern Recognition Letters, 2021Maxafl: Maximizing code coverage with a gradient-based optimization technique.
Youngjoon Kim and Jiwon Yoon
Electronics, 2020
Domestic Patents
DEVICE AND METHOD FOR DATA-ORIENTED ATTACK DETECTION AND RECOVERY FOR ROBOTIC VEHICLES BASED ON SAFETY-CRITICAL VARIABLES MONITORING.
Sangbin Park, Youngjoon Kim, and Donghun Lee
Korean Patent 10-2023-0157140(application number), In reviewSOFTWARE TAINT ANALYSIS METHOD AND SOFTWARE TAINT ANALYSIS DEVICE USING THE SAME.
Kyeongsik Lee, Youngjoon Kim, Younggi Park, and Hojun Lee
Korean Patent 10-2344497-0000, 2021
Other Experiences
| 1-day Vulnerability Analysis | Apr. 2019 – Nov. 2021 |
| Student Intern | Sponsored by Korea University |
| SW Maestro | Jun. 2015 – Dec. 2015 |
| Developer | Sponsored by Ministry of Science and IC |
| Android Malware Anlaysis | Mar. 2014 – Dec. 2014 |
| Student Intern | Sponsored by KISA |
| SGen club | Jul. 2012 – Jun. 2014 |
| Developer | Sponsored by Samsung SDS |
Skills
- Programming Languages: Proficient - C/C++, Python, Java, Occasional - Java, JavaScript, Node.js, R, Solidity, Rust
- Cloud Platforms: AWS, Google Cloud
- Frameworks/Tools: AFL, Pintool, Burp suite, IDA, WinDBG, PyTorch, TensorFlow, React, Git, MongoDB