Core Research Team
The DY-LLM team consists of interdisciplinary experts dedicated to building and optimizing large-scale generative AI models. Our team includes natural language processing specialists, machine learning engineers, data scientists, and algorithm researchers, committed to applying cutting-edge technologies to real-world scenarios like text generation and image synthesis.
Chuanyou Li (Team Lead)
Southeast University · PhD/Professor
Chief Scientist
Research Focus: Artificial Intelligence & Machine Learning
Key Contributions: Published numerous papers in top-tier journals, led fundamental algorithm research for AI large models, specializing in AIGC, high-performance computing acceleration, combinatorial optimization, and LLM system architecture.
Kehua Chen
University of Electronic Science and Technology of China · PhD
AI Agent Development Engineer
Research Focus: Government Information Systems
Key Contributions: Contributed to national key research projects, driving AI innovation in government system optimization and implementation.
Guoxiong Tang
University of Pennsylvania · PhD
Deep Learning Engineer
Research Focus: Machine Learning
Key Contributions: Developed high-efficiency machine learning models and algorithms that significantly improved computational efficiency and accuracy in complex data processing.
Wenxin Zhang
University of Rennes · PhD
LLM Product Director
Research Focus: Modern Management Theory
Key Contributions: Pioneered streaming engineering design management methodologies that became driving forces for regional economic development.
Ningyu Liang
University of Glasgow · MSc
Data Analysis Engineer
Research Focus: Data Mining
Key Contributions: Developed a high-performance data processing system for efficient handling of large-scale heterogeneous data sources.
Sida Xing
Deakin University · MSc
LLM Algorithm Engineer
Research Focus: Recurrent Neural Networks (RNN)
Key Contributions: Developed Enhanced LSTM (E-LSTM) to address gradient issues in traditional RNNs, improving training efficiency across multiple domains.