Skip to content
View ji24077's full-sized avatar
🤓
🤓

Highlights

  • Pro

Block or report ji24077

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
ji24077/README.md

Hi, I’m @ji24077 (Ji Sung Han)! 👋

👀 I'm passionate about various fields, including AI Systems, Machine Learning, Robotics, and Software Engineering, all aimed at solving real-world problems.

🌱 I’m currently studying Computer Science and Statistics (Quantitative Finance) at the University of Toronto.

💡 This spring, I’m working as a Systems Software Engineer Intern at NVIDIA, following a previous internship at Tesla.

I’m working on data-centric post-training for Cosmos3, including weak-reference paired learning, multimodal SFT, and evaluation. I also develop error-injection synthetic-data and evaluation systems for MCP-based computer-use agents.

I am also exploring and contributing to vLLM through an open upstream performance PR focused on inference performance and scheduler efficiency.

📫 Reach out to me via email at jishan.han@mail.utoronto.ca.

🚀 I'm excited to launch my GitHub portfolio soon to showcase and build innovative projects. Stay tuned!

🌍 Check out my personal website: jisunghan.netlify.app – a space where I share my projects, experience, and insights.


About Me 🔍

💻 Bridging AI & Software Engineering
My passion lies in bridging the gap between AI and software engineering, striving to become a versatile developer who excels in both domains. I believe that mastering both AI and software development allows for more impactful and scalable solutions.

📚 Interdisciplinary Thinker
Driven by endless curiosity, I explore diverse subjects beyond my core focus, believing that learning is a lifelong journey. This interdisciplinary approach has shaped my unique perspective on problem-solving and innovation, allowing me to tackle challenges creatively.

🔧 System Programming Enthusiast
I have a strong affinity for system programming, especially with C++ and Linux. The low-level intricacies and fine-grained control it provides fascinate me, reinforcing my appreciation for efficiency and optimization.

🤖 Exploring Robotics, AI Systems & GPU Computing
While my foundation is in Computer Science and Statistics, my curiosity extends into Robotics, GPU Computing, and AI Infrastructure. I’m actively exploring how large-scale AI models can be trained, served, and optimized to power real-world robotics and hardware applications.

🌍 Adaptable & Globally Minded
Having studied abroad and adapted to various environments, I thrive in dynamic settings and excel in making new connections. I strongly believe that technology is a global language, bringing people together to innovate and solve pressing problems.


🚀 Let's Connect!

💼 LinkedIn: https://www.linkedin.com/in/ji24077/

🌍 Personal Website: jisunghan.netlify.app

Let's collaborate and explore the endless possibilities of technology together! 🚀✨

Pinned Loading

  1. AI-Image-Detector AI-Image-Detector Public

    This is final project of Advanced Machine Learning & Deep Learning course at University of Toronto.

    Jupyter Notebook 2

  2. Cerberus-Detection Cerberus-Detection Public

    This is final project of Applied Fundamentals of Deep Learning at University of Toronto

    Jupyter Notebook 1

  3. Machine-Learning-Driven-Quantitative-Trading Machine-Learning-Driven-Quantitative-Trading Public

    This is final project of Advanced Computational Finance course at University of Toronto.

    HTML 2

  4. DreamSpeed DreamSpeed Public

    A from-scratch C++/CUDA LLM inference engine for studying high-performance serving systems.

    C++

  5. UTSC-CSCC09-Programming-on-the-Web/project-meaaji UTSC-CSCC09-Programming-on-the-Web/project-meaaji Public archive

    project-meaaji created by GitHub Classroom

    HTML 1

  6. Reward-Hacking-Resistant-RL-Environment-for-ML-Systems Reward-Hacking-Resistant-RL-Environment-for-ML-Systems Public

    End-to-end RL environment design: sparse MoE task, tamper-resistant judge, and PyTorch performance analysis inspired by ScatterMoE

    Python