Jiheon Choi

Jiheon Choi (최지헌)
Ph.D. candidate, Distributed and Parallel Computing Lab
Department of Artificial Intelligence, Ajou University
Suwon, Republic of Korea
Email: unidev [at] ajou.ac.kr
CV (PDF) | Google Scholar | GitHub | LinkedIn

About Me

Jiheon Choi (최지헌) received a B.S. degree in Software and Computer Engineering from Ajou University in Suwon, South Korea in August 2023. Currently, I am a Ph.D. candidate in the Department of Artificial Intelligence at Ajou University. I am a software and research engineer with over six years of experience in both academic research and industry applications. I am a member of the Distributed and Parallel Computing Lab, led by Ph.D. Sangyoon Oh. My research interests include high-performance computing, distributed AI/ML systems, and application security. You can contact me via email at unidev [at] ajou.ac.kr.

My research has been driven by two primary keywords: HPC and Scalable, Secure AI. My research interests include, but are not limited to:

• High-Performance & Distributed Computing: HPC System Optimization and Job Scheduling

• Systems for AI, AI for Systems: Distributed ML Architecture (i.e., Distributed LLM and GNN), Federated Learning and Privacy-Preserving ML Techniques, Machine Learning for Resource Optimization

• AI for Science: Molecular Dynamics, Machine Learning Interatomic Potential

• Security & Privacy: Privacy-enhancing Techniques (i.e., Differential Privacy and Homomorphic Encryption), Automatic Security Vulnerability Assessment, Secure ML Systems Design

News

• 2026 Aug. Attended Euro-Par 2026 in Pisa, Italy.

• 2026 Aug. Our CCGrid 2026 paper is now available on IEEE Xplore.

• 2026 Jul. Completed all required coursework for the Ph.D. program in Artificial Intelligence at Ajou University.

• 2026 May. S-CQR accepted at Euro-Par 2026.

• 2026 Mar. UARP accepted at The Journal of Supercomputing.

• 2026 Feb. One paper accepted at CCGrid 2026.

• 2025 Sep. One paper accepted at Journal of Systems Architecture.

• 2025 Feb. Presented “When HPC Scheduling Meets Active Learning: Maximizing Performance with Minimal Data” at HPC Asia.

Education

• Ph.D. in Artificial Intelligence (in progress), Ajou University. Advisor: Sangyoon Oh. Sep. 2023 - Present

• B.Sc. in Software and Computer Engineering, Ajou University. Aug. 2023

Publications [Google Scholar]

2026

• [C3] S-CQR: Stratified Calibration for Runtime Prediction in HPC Backfill Scheduling. Jiheon Choi, Sangyoon Oh. 32nd International European Conference on Parallel and Distributed Computing (Euro-Par), 2026. [DOI]

• [C2] Reducing Backfill Failures from Workload Drift with Lightweight Uncertainty Buffers in HPC Job Scheduling. Jiheon Choi, Sangyoon Oh. 26th IEEE International Symposium on Cluster, Cloud and Internet Computing (CCGrid), 2026. [DOI]

• [J3] UARP: Uncertainty-Aware Runtime Prediction for Preventing Scheduler Termination under Wallclock Constraints in HPC. Jiheon Choi, Sangyoon Oh. The Journal of Supercomputing, 2026. [DOI]

2025

• [J2] Lightweight Multi-Layered De-Identification Architecture: Secure Client Selection in Federated Learning. Jiheon Choi, Sangyoon Oh. Journal of Systems Architecture, 2025. [DOI]

• [C1] When HPC Scheduling Meets Active Learning: Maximizing the Performance with Minimal Data. Jiheon Choi, Jaehyun Lee, Minsol Choo, Taeyoung Yoon, Oh-Kyoung Kwon, Sangyoon Oh. High Performance Computing in Asia-Pacific Region (HPC Asia), 2025. [DOI]

2024

• [J1] Staleness Aware Semi-asynchronous Federated Learning. Miri Yu, Jiheon Choi, Jaehyun Lee, Sangyoon Oh. Journal of Parallel and Distributed Computing, 2024. [DOI]

See also: Domestic publications (English and Korean).

Patents

• Task Scheduling Device, Task Scheduling method thereof and Computing System. South Korea 10-2025-0078984 (Applied)

• Method and Apparatus for Classifying Workload Using Artificial Intelligence Model. South Korea 10-2023-0029398 (Applied)

• Apparatus and Method for Hotspot Prevention Based on Bloom Filter in Distributed Database Management System. South Korea 10-2024-0162896 (Applied)

• System and Method for Detecting Dynamic Web Vulnerability of Data Normalization of HTTP Headers for Classification of HTTP Headers. South Korea 10-2024-0159333 (Applied)

• Method for Automatic Diagnosis Vulnerability of Web and Apparatus for Performing The Method. South Korea 10-2311119 (Registered)

• System and Method for Analyzing Malicious Code Based on Blockchain. South Korea 10-222804 (Registered)

Work and Research Experience

• Graduate Research Assistant, Distributed and Parallel Computing Lab, Ajou University, Suwon, Republic of Korea (Sep. 2023 - Present)

  • HPC Scheduling: Developed active-learning runtime predictors, uncertainty-aware prediction, adaptive buffers for workload drift, and stratified calibration [C1, J3, C2, C3]. S-CQR reduced mean wait time by 1.2-28.3% versus UARP in EASY backfill simulations on five ALCF traces, with job kill rates below 1%.

  • LLM Inference Systems: Profiled vector-indexing bottlenecks in high-concurrency retrieval-augmented generation (RAG) workloads and redesigned storage access using a multi-tier hierarchy for LLM serving.

  • GPU Graph Processing: Developed memory-efficient graph partitioning and fine-grained GPU kernels for billion-scale graph workloads and GNN inference in molecular dynamics simulations.

  • Privacy-Preserving Machine Learning: Developed a lightweight multi-layered de-identification architecture for secure client selection and studied staleness-aware aggregation in semi-asynchronous federated learning [J1, J2].

• Lead Software and Security Research Engineer, MONTHLY HACKING Inc., Seoul, Republic of Korea (Aug. 2020 - Dec. 2023)

  • Delivered security consulting to 60+ domestic startups, including penetration testing, vulnerability assessment, and incident response during security breaches.

  • Built backend services for an automated penetration-testing SaaS platform and led development of a web vulnerability scanner that passed a government-certified performance evaluation for an SME R&D grant.

  • Automated Log4j vulnerability detection and reporting; integrated the scanning engine into a Spring Boot production backend through a Python Flask API.

  • Developed static analysis for vulnerabilities in mobile application binaries and delivered it as a B2B SaaS product using Nest.js and TypeScript.

• Software Engineer, WAPAS SYSTEMS Inc., Incheon, Republic of Korea (Oct 2019 - Aug 2020)

  • Built a mobile web ordering and delivery platform for the Philippine market using Node.js, React, and AWS, including customer and administrative workflows.

  • Designed cash-based settlement for local payment infrastructure, supporting household and enterprise order-to-delivery workflows where credit card adoption was limited.

  • Researched mobile OS security architectures on Android and iOS, developing bypass methods that supported a new funded project.

• Other Experience

  • BSKB's Summer Patent Seminar, Birch, Stewart, Kolasch & Birch, LLP (Jun 2023): Training in U.S. patent prosecution, litigation strategy, recent developments, and case studies.

Research Projects

Research team participation; dates indicate project periods. Future end dates are planned.

• [National Research Foundation (NRF)] Efficient Hyperscale LLM Inference System based on Scale-out Context Memory. Jul. 2026 - Jun. 2029 (planned)

• [KEIT (funder) · HL Mando (industry partner)] Cloud-updatable SDV chassis software for heterogeneous zonal control units (ZCUs). Apr. 2026 - Dec. 2029 (planned)

  • Investigate federated learning applications for autonomous, software-defined vehicles.

• [KISTI · National Research Council of Science and Technology] A Study on Graph Parallel Processing Algorithm Based on GPU. Jun. 2024 - Present

  • GPU acceleration of molecular dynamics simulations using graph neural networks.

• [Ministry of Science and ICT] Development of multi-tiered, multi-purpose autotuning framework for exa-scale supercomputers. Nov. 2023 - Present

• [Samsung Display Inc.] High-efficiency HPC job scheduling algorithms. Improved utilization from 72% to 93% in simulations using job logs. Jan. 2024 - Feb. 2025

• [Korea Automotive Technology Institute] Safety of the Intended Functionality (SOTIF) for perception and decision-making insufficiency. Sep. 2022 - Present

• [IITP] MR-IoT convergence AI for disaster countermeasures. Jan. 2023 - Dec. 2023

• [Ajou University] User engagement in social/mobile platforms and welfare value. Jan. 2022 - Oct. 2022

  • Provided web crawling and scraping for social engagement data, including HackerOne and Bugcrowd.

• [Ministry of SMEs and Startups] Web application vulnerability assessment software. Apr. 2022 - Dec. 2022

Technical Skills

• Programming Languages: Python, C, C++, Java, JavaScript, TypeScript

• HPC and Cluster Management: SLURM, IBM LSF, MPI

• Machine Learning and LLM Systems: PyTorch, vLLM, DeepSpeed, Hugging Face, LangChain

• Web and Backend Systems: Node.js, React, Next.js, Nest.js, Spring Boot, Flask, AWS

• Security Tools: Burp Suite, Nmap, Wireshark, Frida

• Data Analysis and Visualization: Pandas, Matplotlib

Professional Services

• Reviewer

  • IEEE Transactions on Parallel and Distributed Systems (2026)

  • IEEE Transactions on Vehicular Technology (2026)

  • Nature Scientific Reports (2026)

  • Cluster Computing (2026)

  • Discover Artificial Intelligence (2026)

  • Journal of Network and Systems Management (2026)

  • International Journal of Machine Learning and Cybernetics (2026)

• Teaching Assistant, Ajou University. Jan. 2022 - Present

  • Operating Systems: Fall 2026

  • Distributed and Parallel Computing: Fall 2025

  • Object-oriented Programming: Spring 2024, Spring 2025, Spring 2026

  • Web System Design: Fall 2023, Fall 2024

  • Artificial Intelligence Convergence Capstone Design: Fall 2022, Spring 2023

  • Data Analytics-Machine Learning: Spring 2022

• Mentor, Developer Bootcamp Mentoring

  • Team Sparta Inc. (Jul 2022 - May 2023): Mentored learners in modern web frameworks, including React and Node.js.

  • Elice Inc. (Nov 2021 - Aug 2022): Mentored learners in modern web frameworks, including React and Node.js.

• Mentor, Korea Open Source Software Association (Oct 2020 - Dec 2020)

  • Mentored participants in the Open Source Software Developer Competition hosted by the Ministry of Science and ICT.

• Instructor, Korea Information Technology Research Institute (Aug 2021 - Dec 2021)

  • Taught ethical hacking and web, mobile, and system security in the Cyber Guardians program for high school students.