Letian Ruan

Hello! I'm a junior student at University of Michigan and Shanghai Jiao Tong University pursuing dual Bachelor's degrees.

Currently, I'm doing research at Catalyst Group in Carnegie Mellon University, advised by Prof. Zhihao Jia, and SymbioticLab in University of Michigan, advised by Prof. Mosharaf Chowdhury. Previously, I was fortunate to work with Prof. Shixuan Sun at EPCC Lab in Shanghai Jiao Tong University. I have interned at MiniMax and was a proud member of SGLang and Mooncake.

My research interests mainly lie in Distributed and Machine Learning Systems, especially in Serving Systems (Agentic/Robotics/Multimodal), ML Compiler and RL Infra.

Email / CV / Google Scholar / Github

News

[2026.04] Glad to share our recent work in SGLang-RL team to optimize refitting for large-scale RL training. Feel free to check out our blog on LMSYS.Org!
[2026.03] I'm attending ASPLOS2026 in Pittsburgh. Feel free to reach out!
[2026.02] Excited to announce the release of Forge, a scalable Agent RL framework powering the M2~M3 series models.
[2026.01] Our paper of IDP is accepted by EuroSys 2026! Congradulations!
[2025.12] Our work FaaSBoard is accepted by SIGMOD 2026, check out the paper!

more previous news...

Publications

	InfiniLoRA: Disaggregated Multi-LoRA Serving for Large Language Models Hongyu Chen, Letian Ruan, Zilin Xu, Yuchen Li, Xinyu Chen, Jingwen Leng, Bingsheng He, Minyi Guo, Shixuan Sun Arxiv, 2026 Preprint / Github
	Bridging the GPU Utilization Gap: Predictive Multi-Dimensional Resource Scheduling for AI Workloads Yilei Lu, Dongbiao He, Teng Ma, Zhe Liu, Letian Ruan, Jinlei Jiang, Yongwei Wu. EuroSys, 2026 Published Paper / Github
	FaaSBoard: Efficient Graph Processing with a Disaggregated Architecture on Serverless Services Yushi Liu, Yikang Ruan, Letian Ruan, Zijun Li, Sen Gao, Weihao Cui, Shixuan Sun, Quan Chen, Shuo Quan, Jie Wu, Bingsheng He, Minyi Guo SIGMOD, 2026 Published Paper / Github

Blogs

	Updating 1T parameters in seconds — P2P weight transfer in Large Scale Distributed RL Jiadong Guo, Xin Ji, Letian Ruan, Teng Ma, Chenyang Zhao, Yueming Yuan, Zhichen Zeng SGLang-RL Team, LMSYS.Org April 2026, Post
	Forge: Scalable Agent RL Framework and Algorithm MiniMax Team Published alongside the MiniMax M2.5 Tech Report Feb. 2026, Tech Report

Experience

	Carnegie Mellon University 2026.04 ~ Present Pittsburgh, PA, USA Optimize Megakernel Compilers and Agentic Serving Systems. Advisor: Prof. Zhihao Jia Visiting Student Researcher, Catalyst Group
	MiniMax 2025.12 ~ 2026.02 Shanghai, China Work on the large-scale Agent RL framework that powered breakthrough capabilities in the M2~M3 series models. Supervisor: Yuelan System Software intern, RL Infra team
	University of Michigan, Ann Arbor 2025.08 ~ Present Ann Arbor, MI, USA Promote the Robotics Serving Systems and Any-to-Any Multimodal Models Serving Systems. Advisor: Prof. Mosharaf Chowdhury. B.S.E. in Computer Science, CSE Department
	Shanghai Jiao Tong University 2022.09 ~ Present Shanghai, China Reduce the long-tail latency for Multi-LoRA serving and build the disaggregated arch for serverless graph processing. Advisor: Prof. Shixuan Sun B.S. in Mechanical Engineering (duel degree), Global College

Misc.

I'm a sport enthusiast, especially in basketball and running. I've been watching NBA since 2018, and a big fan of the Golden State Warriors and Stephen Curry!
Hollow Knight is my favorite video game, which took me about 50 hours to complete whole challenges.