Hi, there!
I’m Jiufeng (pronounced as /jo phone/) from China, you can also call me FENG (/phone/)! Welcome to my personal website. :D
I’m currently an ML Engineer at Apple, working on ML/AI, LLMs, system performance analysis, and AI Agents. Prior to Apple, I completed my master’s degree in Data & Computer Science at Heidelberg University and finished my Data Science & AI internship and working student position at Mercedes-Benz Group on Recommendation Systems and AI Agents. I’m deeply passionate about ML research, AI Agents, reinforcement learning (especially Agentic RL), and the full lifecycle of large language models — from pretraining and posttraining to inference optimization. I’m also excited about multimodal AI and its potential to reshape how we interact with intelligent systems.
ML/AI Engineer & Researcher with 3+ years building machine learning systems across academic research and production environments. Focused on LLM pretraining & posttraining, reinforcement learning, AI Agents, and inference optimization. Proven track record in NLP/RAG (Uni Heidelberg & NLP Laboratory), production recommender systems (Mercedes-Benz Group), computer vision, and multimodal AI. Specialized in neural network optimization with expertise spanning both research and engineering: efficient quantization strategies, mixed-precision inference optimization (Uni Heidelberg & HAWAII Lab), and GPU-accelerated training – delivering production-grade low-latency inference systems with minimal accuracy loss.
Outside of tech, I’m passionate about art, music, and film. I love watching films, and I have many beloved movie directors like Kar Wai Wong, Edward Yang, Alfred Hitchcock, Hirokazu Kore-eda, Andrei Tarkovsky and many others. Music is also close to my heart; I enjoy everything from classical to jazz, hip-hop, and rock & roll. Frédéric Chopin and Johann Sebastian Bach are two of my all-time favorites, and their music has been a constant companion in my life.
When I’m not coding or working on projects, you’ll often find me reading and doing sports — these are my favorite ways to unwind. Fitness is a big part of my life too; I’ve been a gym rat for almost three years now, and I also enjoy swimming, playing basketball, and hiking.
Education
Heidelberg University (DE)
M.Sc. Data and Computer Science | October 2022 - August 2025
Graduated with a master’s degree focused on advanced machine learning, neural network optimization, and AI systems. My thesis research centered on “Mix-Precision Stochastic Quantization of Activations during Backpropagation” — a cutting-edge approach to optimizing deep neural networks for improved efficiency and performance.
Key Coursework: Advanced Machine Learning, Deep Learning, Natural Language Processing, Computer Vision, High-Performance Computing, GPU Computing, Computational Geometry, Foundations of LLMs and Generative AI.
Northeastern University (CN)
B.Sc. Computer Science | September 2016 - June 2020
Graduated from a prestigious 211 & 985 project university (ESI Rank Top 1‰ in Computer Science). My undergraduate thesis focused on “Construction of a Multimodal Knowledge Graph for Fine-Grained Sentiment Analysis” — pioneering work in combining multiple data modalities for enhanced AI understanding. → Emotion Net
Academic Achievements:
- China Invention Patent (Granted): A Method for Constructing and Presenting a Multimodal Sentiment Knowledge Graph — CN ZL202011319237.8, granted September 7, 2021; co-inventor
- The Second Prize Scholarship 2018 & 2019(NEU, CN)
- Outstanding Graduation Thesis 2020(NEU, CN)
Research & Technical Expertise
Core Areas of Expertise
- LLM Pretraining & Posttraining: Large-scale model training, RLHF/RLAIF alignment, and fine-tuning strategies
- AI Agents & Agentic RL: Autonomous agent architectures, tool use, and reinforcement learning for agentic systems
- Inference Optimization: Mixed-precision quantization, GPU-accelerated inference, and low-latency serving at scale
- Multimodal AI: Vision-language models, cross-modal learning, and multimodal knowledge representation
- Production ML Systems: End-to-end recommender systems, RAG pipelines, and MLOps infrastructure
Research Contributions
- Master’s Thesis: Advancing the field of neural network quantization with novel stochastic approaches for activation quantization during backpropagation
- Patent Holder: Innovative multimodal AI systems for sentiment analysis and knowledge graph construction
- Industry Impact: Deployed AI systems processing terabytes of data daily, serving millions of users across multiple markets
Technical Skills
- Languages: Python, C++, SQL, JavaScript
- ML/DL Frameworks: PyTorch, JAX, Hugging Face, LangChain, vLLM
- Cloud & Infrastructure: AWS, Azure, Databricks, Docker, Kubernetes
- RL & Alignment: PPO, GRPO, DPO, RLHF, reward modeling
- Specializations: LLMs/GenAI, AI Agents, Reinforcement Learning, Inference Optimization, Multimodal AI
Tags on me:
ENTJ | Gym | AI/ML | Movies | Music | Reading | Writing | Travelling | Photograph | Cooking | Fashion | Entrepreneurship
Research Interests
- AI Agents & Agentic RL: Building autonomous agents that learn through interaction, with reinforcement learning driving adaptive decision-making and tool use
- LLM Pretraining & Posttraining: Scaling laws, data curation, alignment techniques (RLHF/DPO), and improving model capabilities through training methodology
- Inference Optimization: Quantization, speculative decoding, KV-cache optimization, and hardware-aware serving for efficient deployment
- Multimodal AI: Vision-language models, cross-modal reasoning, and unified architectures for understanding diverse data modalities
- Reinforcement Learning: Reward modeling, policy optimization, and RL-driven approaches to training and aligning large models
Hobbies
- Film Enthusiast: Deeply passionate about cinema, especially works by auteur directors like Kar Wai Wong, Edward Yang, and Alfred Hitchcock
- Music Lover: Eclectic taste spanning classical, jazz, hip-hop, and rock & roll — with Chopin and Bach as eternal favorites
- Avid Reader: Always have a book in hand, from research papers and technical reports on LLMs to literature, business insights, and philosophy
- Gym Rat: Dedicated gym-goer for three years, also enjoy swimming, basketball, ping pong, and hiking
- Travel & Photography: Love exploring new places with my camera, capturing moments and cultures
Goals & Aspirations
- Professional: Leverage my advanced AI education to drive innovation and become a leading expert in AI optimization
- Entrepreneurial: Launch an AI startup that creates real value for people and pushes the boundaries of what’s possible
- Personal: Travel to 25+ countries, experience diverse cultures, and continue learning from the world
- Impact: Help shape the future of artificial intelligence and contribute to meaningful technological advancement
Connect With Me
- Email: lijiufeng97 at gmail dot com
- LinkedIn: Jiufeng Li
- X: @jiufeng0924
- GitHub: CoreSheep
- Instagram: @jiufeng.li
“The only way to do great work is to love what you do. If you haven’t found it yet, keep looking. Don’t settle.”
— Steve Jobs