Scholar
Xianzhong Ding
Google Scholar ID: lUNyhjwAAAAJ
Accenture
Citations & Impact
Citations
673
H-index
13
i10-index
14
Publications
20
Co-authors
0
Publications
Resume (English only)
Academic Achievements
- - 2025/02 Invited as Guest Editor for IoT 2025.
- - 2025/02 One paper got accepted to IEEE IoT-J 2025.
- - 2025/01 Won the Best Paper Award at HICSS 2025.
- - 2024/12 One paper got accepted to EuroSys 2025.
- - [EuroSys'25] Towards VM Rescheduling Optimization Through Deep Reinforcement Learning
- - [IoT-J'25] A Safe and Data-efficient Model-based Reinforcement Learning System for HVAC Control
- - [HICSS'25] Deepot: Parking Lot Identification Using Low-Resolution Satellite Imagery, Best Paper Award
- - [TOSN'24] Optimizing Irrigation Efficiency Using Deep Reinforcement Learning in the Field
Research Experience
- - Advanced AI Research Scientist at Accenture, focusing on multi-agent systems, LLM orchestration, and scalable AI infrastructure.
- - Postdoctoral Researcher at Lawrence Berkeley National Laboratory (LBNL), developed play-verl — a VERL-based reinforcement learning benchmark evaluating PPO and GRPO algorithms on Qwen models with distributed training on multi-GPU systems; also worked on LLM fine-tuning for freight infrastructure and large-scale EV simulations.
Education
- Ph.D. in Computer Science and Engineering, University of California, Merced (UCM).
Background
- Research Interests: Large Language Models (LLMs) and Multi-Agent Systems, Reinforcement Learning and Reinforcement Learning from Human Feedback (RLHF), Distributed Training and Scalable AI Infrastructure, Machine Learning for Systems and Resource Optimization.
Miscellany
- Technical Skills: Python, C/C++, SQL; PyTorch, JAX, LangChain, AutoGen, FastAPI, Ray, DeepSpeed, FSDP; Multi-Agent Systems, RAG/GraphRAG, MCP, RLHF, VERL, LLMOps; Azure, AWS, Kubernetes, Docker, Redis, PostgreSQL, Databricks, Snowflake; Git, GitHub Actions, Helm, pytest, Black, Pylint, Pyright.
Co-authors