Kai Shen | University of Toronto

Short Bio

I am a third-year Ph.D. candidate in the Department of Electrical and Computer Engineering at University of Toronto, where I am co-supervised by Prof. Angela Demke Brown and Prof. Eyal de Lara as part of the Computer System & Networks Group. I received my B.Eng. degree with Honours, First Class from The Chinese University of Hong Kong, Shenzhen in 2022, majored in Computer Science and Engineering under the supervision of Prof. Fangxin Wang.

My research lies broadly in computer systems, with a focus on memory management across the stack, from the OS kernel to AI serving infrastructure. I am currently working on:

Tiered memory systems: runtime-aware page placement and reclamation across DRAM and CXL memory, leveraging garbage-collection insights from managed runtimes.
AI infrastructure: KV cache management for agentic LLM serving, deciding what to keep, offload, and evict across GPU and CPU memory.

From 2023 to early 2024, I developed packet-level, Transformer-based network performance estimators, as well as a Rust-powered, process-based high-performance network simulator. Earlier, my work focused on server-driven adaptive realtime video streaming.

Research Interests

Operating Systems: memory management, tiered memory
AI Infrastructure: LLM serving, KV cache management
Computer Networks: network simulation, multimedia streaming

Publications

TMC'24

TrimStream: Adaptive Realtime Video Streaming Through Intelligent Frame Retrospection in Adverse Network Conditions

Dayou Zhang, Lai Wei, Kai Shen, Hao Zhu, Dan Wang, Fangxin Wang

IEEE Transactions on Mobile Computing (TMC), 2024.

PDF
TMC'23

DSJA: Distributed Server-Driven Joint Route Scheduling and Streaming Adaptation for Multi-Party Realtime Video Streaming

Dayou Zhang, Hao Zhu, Kai Shen, Dan Wang, Fangxin Wang

IEEE Transactions on Mobile Computing (TMC), 2023.

PDF
IEEE Network'23

Learning-Based Network Performance Estimators: The Next Frontier for Network Simulation

Kai Shen, Baochun Li

IEEE Network, Special Issue on Interplay between Machine Learning and Networking Systems (IEEE Network), 2023.

PDF
INFOCOM'23

SJA: Server-driven Joint Adaptation of Loss and Bitrate for Multi-Party Realtime Video Streaming

Kai Shen, Dayou Zhang, Zi Zhu, Lei Zhang, Fangxin Wang, Dan Wang

IEEE International Conference on Computer Communications (INFOCOM), 2023.

PDF
ICME'22

Towards Joint Loss and Bitrate Adaptation in Realtime Video Streaming

Dayou Zhang*, Kai Shen*, Fangxin Wang, Dan Wang, Jiangchuan Liu (* Co-First Authors)

IEEE International Conference on Multimedia and Expo (ICME), 2022.

PDF

Services

Conference Reviewers

ACM Multimedia 2024

Journal Reviewers

Teaching

Teaching Assistant at The Chinese University of Hong Kong, Shenzhen

CSC3150: Operating Systems, Spring'22

Teaching Assistant at University of Toronto

CSC148: Introduction to Computer Science, Winter'23
APS105: Computer Fundamentals, Winter'24 and Summer'24
CSCC43: Introduction to Databases, Summer'24
CSC369: Operating Systems, Fall'23 and Fall'24
ECE361: Computer Networks, Fall'24

Awards

The Edward S. Rogers Sr. Graduate Scholarships, University of Toronto, 2022-2024
School of Data Science Dean's List Award, CUHK(SZ), 2019-2022
The 17th and 18th Round Undergraduate Research Awards, CUHK(SZ), 2021-2022
Academic Performance Scholarship, CUHK(SZ), 2020-2021
Bowen Scholarship, CUHK(SZ), 2018-2022

Contact

Address: 40 St George St, Toronto, ON M5S 2E4
Office Location: Bahen Centre for Information Technology 5214
Email: kai (at) cs.toronto.edu

Webpage credits