Letian Ruan

 |  News  |  Experience  |  Publications  |  Blogs  |  Miscellaneous  | 

Hi, I'm Letian Ruan, a junior student pursuring dual Bacholr's degree at University of Michigan and Shanghai Jiao Tong University.

Now, I'm doing research at SymbioticLab , advised by Mosharaf Chowdhury. Besides, I'm also a proud member of SGLang team from LMSYS.Org and Mooncake project from MADSys Group. During my sophomore year, I was fortunate to work with Shixuan Sun at Emerging Parallel Computing Center.

My research interests mainly lie in Large Model Systems, particularly in disaggregating and saturating hardware resources. Currently, I'm working on RL Infra and VLA Inference. The goal of my research is to push the boundary of how large-scale models can be designed, trained and efficiently served to make physical impact.

~ CV  |  Email  |  GitHub  |  LinkedIn  |  Twitter ~

profile photo

Dec '25  

 Fortunate to receive an internship offer from Minimax to work on RL training infra.

Oct '25  

 Glad to share that I've joined the SGLang team as a tech staff.

Aug '25  

 Happy to become part of SymbioticLab , advised by Mosharaf Chowdhury .

May '25  

 Start my internship at KVCache.AI as a research intern, mentored by Teng Ma and Mingxing Zhang.

Mar '25  

 Got admitted to the University of Michigan through dual bachelor's degree program.

Oct '24  

 Joined the Emerging Parallel Computing Center as a research assistant, advised by Shixuan Sun
Undergraduate Student | University of Michigan, Ann Arbor
August '25 - Present

Serve Any-to-Any Multimodal LLMs and accelerate VLA model inference.
Advisor: Mosharaf Chowdhury at SymbioticLab.

Undergraduate Student | Shanghai Jiao Tong University
September '22 - Present

Reduce latency for Multi-LoRA serving and build disaggregated arch for serverless graph processing.
Advisor: Shixuan Sun at Emerging Parallel Computing Center.

Development Intern | Minimax
December '25 - Present

RL Infra team, responsible for developing frameworks for training production models.

Tech Member | SGLang
October '25 - Present

Member of SGLang team, responsible for supporting RL features.

Research Intern | KVCache-AI.Org
May '25 - Nov '25

Core dev of Mooncake from MADSys Group at THU.
Promote disaggregated storage and communication layers for large model systems.
Mentor: Mingxing Zhang and Teng Ma.

* means equal contribution here

 Bridging the GPU Utilization Gap: Predictive Multi-Dimensional Resource Scheduling for AI Workloads

[paper] [code]

under review

FaaSBoard: Efficient Graph Processing with a Disaggregated Architecture on Serverless Services

[paper] [code]

under review

  • More blogs coming soon.
  • 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.
 

This template is a modification to Jon Barron's website.