Hailong Jiang
he/him/his
Assistant Professor
Computer Science and Information System
Youngstown State University
Biography
Dr. Hailong Jiang is an Assistant Professor of Computer Science and Information Systems at Youngstown State University (YSU). His research centers on high-performance computing (HPC) resilience, compiler intermediate representations (LLVM and MLIR), and the integration of large language models (LLMs) for intelligent compiler analysis and optimization. Before joining YSU, he completed his Ph.D. in Computer Science at Kent State University, where he developed IR-level fault tolerance mechanisms and LLM-assisted code reasoning frameworks. Dr. Jiang currently leads several initiatives to build open-source datasets and tools that connect HPC compiler research with practical educational applications. His ongoing projectsâsuch as LLM4IR and AAPAsetâfocus on leveraging AI to discover optimization opportunities that traditional compilers miss, particularly in GPU and vectorization pipelines. At YSU, he actively mentors undergraduate and graduate students in system-level programming, data structures, and operating systems, promoting project-based learning aligned with ABET outcomes. His broader goal is to create sustainable research pathways linking regional institutions to national laboratories. Through SRP, he aims to collaborate with DOE and NSF researchers on AI-driven compiler intelligence and performance-portable HPC workflows, providing students with authentic research experiences in scalable computing environments.
Degrees Earned
Ph.D. in Computer Science Kent State University, Kent, OH August 2018 to August 2025 Dissertation: "Research on resilience in high-performance Computing (HPC) applications with Large Language Models" M.S. in IC Engineering University of Chinese Academy of Science, Beijing, China August 2014 to May 2017 Thesis: âThe Study of Cu2ZnSnS4 films generation by sulfur-free annealing process and device applicationâ B.S. in Electronic Science and Technology Xidian University, Xiâan, China August 2010 to June 2014 Thesis: âA novel infrared object tracking algorithmâ
Research Areas
Computer Science; Machine Learning/AI
Research Interests
My research focuses on the intersection of compiler optimization, high-performance computing (HPC), and artificial intelligence. I study how compiler intermediate representations (LLVM and MLIR) can be enhanced with large language models (LLMs) to improve code optimization, program analysis, and performance portability across architectures. My current work develops hybrid quantum and AI-assisted compiler frameworks that automatically identify optimization opportunitiesâespecially those missed by traditional static analyses, such as vectorization and GPU parallelization patterns. I am also interested in building reproducible datasets and open-source tools for IR-aware learning, enabling the research community to evaluate LLMs for compiler reasoning tasks. At Youngstown State University, I integrate these research topics into advanced systems and programming courses, mentoring undergraduate and graduate students in project-based research that connects compiler theory, machine learning, and HPC practice. My long-term goal is to create a sustainable research pathway that links regional universities with national laboratories to advance intelligent compiler systems and scalable scientific computing.
Topical Areas
Computer Science
Research Synergy
My research naturally aligns with the mission of DOE and NSF laboratories to advance scalable, intelligent, and sustainable computing. The intersection of compiler optimization, AI, and high-performance computing offers fertile ground for collaboration with research groups developing next-generation programming models, runtime systems, and performance-portable workflows. My ongoing work on LLM-assisted compiler analysis and IR-level optimization can directly complement efforts in automatic performance tuning, heterogeneous code generation, and AI-for-Science initiatives within NAIRR and DOEâs Exascale Computing Project. By integrating large language models with traditional compiler infrastructures such as LLVM and MLIR, my team seeks to build interpretable AI agents that reason about code transformations and improve resilience and efficiency across architectures. Through SRP, I aim to collaborate with laboratory scientists to benchmark, validate, and extend these models using real scientific workloads and HPC benchmarks. Such collaborations would accelerate both compiler intelligence research and practical applications in data-intensive and simulation-driven computing, while providing students hands-on experience in scalable and trustworthy AI systems.
Motivation
I want to participate in the Sustainable Research Pathways program because I see it as an opportunity to build lasting collaborations that connect research, education, and community impact. As a new faculty member at Youngstown State University, I am eager to establish partnerships that help my students experience research beyond the classroom and engage with national laboratories working on high-performance and AI-driven computing. My motivation also comes from my own academic journey. I benefited from mentors who opened doors to research computing when I was a student, and I want to create similar opportunities for my students, especially those from institutions with fewer resources. I believe SRPâs model of mentorship and collaboration fits perfectly with this goal. Through this program, I hope to contribute my experience in compiler and AI research, learn from scientists working on scalable systems, and help my students grow as independent researchers who see science as both accessible and collaborative.
Supervising Students Plan
I plan to include my current masterâs teaching assistant as the student member of my SRP team. As part of his graduate training at Youngstown State University, he has been assisting in systems and programming courses while developing strong skills in C++, data structures, and compiler fundamentals. My supervision plan will combine structured weekly mentoring with hands-on project milestones. Before the summer collaboration, we will meet regularly to review foundational materials on LLVM, MLIR, and high-performance computing. During the SRP program, we will coordinate closely with the host lab mentor to align our tasks with ongoing research goals. I will emphasize iterative progress tracking, code documentation, and technical communication to ensure a productive and educational research experience. This approach will allow my TA to contribute meaningfully to the project while developing advanced research and teamwork skills in a real-world HPC environment.
Student Merit
The student I plan to include is a masterâs candidate in Computer Science at Youngstown State University and currently serves as my teaching assistant for upper-level programming and systems courses. His performance in both coursework and teaching duties demonstrates strong technical aptitude, professionalism, and leadership. He consistently shows initiative in helping undergraduate students understand complex programming concepts and has expressed strong interest in pursuing research related to compiler optimization and AI-driven computing. His combination of teaching experience, programming proficiency, and motivation to learn makes him an ideal candidate for the SRP program. He is well prepared to take on research challenges, adapt to collaborative lab environments, and apply his computational knowledge to real scientific workloads. I believe this experience will significantly strengthen his research potential and open pathways for future graduate study or national lab engagement.
Lightning Talk Title
LLM4IR: Large Language Models for Compiler IR Optimization
Keywords
Large Language Model, LLVM IR, MLIR, HPC, Compiler Optimization