Any Questions?
Ask my chatbot, it is built around my profile and has access to my latest updates.
Hi, I’m Seyed Mohammad Hejazi Hoseini, a Research Assistant and Master of Applied Science student in Electrical & Computer Engineering at York University in Toronto.
My research focuses on AI reproducibility, LLM inference, and software engineering agents. I am particularly interested in improving the reliability, traceability, and repeatability of large language model systems in practical applications.
My recent work explores reproducibility challenges in modern large language model inference, including evaluation pipelines, retrieval-augmented generation, and agentic workflows, where small sources of nondeterminism can propagate into significantly different outcomes.
Research Interests
- AI reproducibility
- LLM inference
- Software engineering agents
- Software engineering for AI
- AI for software engineering
Experience
Full details »Research Assistant
Lassonde School of Engineering, York University
Jan. 2025 - Present · Toronto, ON
- Developing a reproducibility framework for inference engines such as llama.cpp and vLLM.
- Error analysis and cost prediction for software engineering agents.
Teaching Assistant
York University
Jan. 2025 - Present
Teaching Assistant
Amirkabir University of Technology - Tehran Polytechnic
Feb. 2022 - Jun. 2022
Education
Full details »York University
Jan. 2025 – Present
MASc in Electrical & Computer Engineering
Toronto, Canada
Amirkabir University of Technology
Sept. 2018 – Jul. 2023
BSc in Computer Engineering
Tehran, Iran
Shahid Ghoddusi High School (NODET)
Sept. 2014 – Jul. 2018
Diploma in Mathematics and Physics
Qom, Iran
Recent News
All news »- [2026/03] Updated my personal website and built a personal profile chatbot.
- [2026/03] Our paper "A record replay framework for reproducible LLM inference under fixed execution conditions, with backend-level RNG capture and replay for llama.cpp and vLLM" is at its final stages of being published!