About Me

I am a postdoctoral researcher in the Systems Group at ETH Zurich, where I am fortunate to work with Dr. Ana Klimović. I received my Ph.D. in Computer Science from Emory University, advised by Dr. Ymir Vigfusson.

My research aims to make distributed systems more efficient and reliable, with a focus on caching, performance diagnosis, and cloud infrastructure. I investigate how these systems respond to changing workloads and failures, from controlling retries in microservices to diagnosing LLM serving failures and adapting web caches to AI-generated traffic.

I am looking for research and engineering opportunities in industry, particularly in distributed systems and AI infrastructure.

Research Focus

AI Serving & Model Deployment (ongoing)
- Model onboarding, serving correctness, and hardware/runtime compatibility
- Agent-assisted diagnosis, repair, and optimization of LLM deployments
Reliable Distributed Systems & Performance Diagnosis
- Retry admission and control in service meshes; retry storms and metastable failures [NSDI’27]
- Distributed tracing and causal modeling of end-to-end latency [SoCC’23]
- Elastic cloud infrastructure, scheduling, and sandboxed execution [SOSP’25]
Caching & Data Infrastructure
- Simple, efficient cache eviction [NSDI’24 🏆] [SOSP’23] [HotOS’23]
- Workload characterization and web cache design for AI-generated traffic [SoCC’25]

Publications

(* == equal contribution)

[NSDI] It’s Time to Retry: Taming Retry Storms by Treating Retry Capacity as a Shared Resource
Yazhuo Zhang, Giuseppe Steduto, Luca Renna, Ana Klimovic
USENIX Symposium on Networked Systems Design and Implementation (NSDI), 2027. Accepted.
[code]

[SoCC] Rethinking Web Cache Design for the AI Era
Yazhuo Zhang, Jinqing Cai, Avani Wildani, Ana Klimovic
The 16th edition of the annual ACM Symposium on Cloud Computing (SoCC), 2025.

[SOSP] Unlocking True Elasticity for the Cloud-Native Era with Dandelion
Tom Kuchler, Pinghe Li, Yazhuo Zhang, Lazar Cvetković, Boris Goranov, Tobias Stocker, Leon Thomm, Simone Kalbermatter, Tim Notter, Andrea Lattuada, Ana Klimovic
The 31th ACM Symposium on Operating Systems Principles (SOSP), 2025
[code]

[NSDI] SIEVE is Simpler than LRU: an Efficient Turn-Key Eviction Algorithm for Web Caches
Yazhuo Zhang*, Juncheng Yang*, Yao Yue, Ymir Vigfusson, K. V. Rashmi.
21st USENIX Symposium on Networked Systems Design and Implementation (NSDI), 2024
[website][blog]
🏆 Community Award

[SoCC] LatenSeer: Causal Modeling of End-to-End Latency Distributions by Harnessing Distributed Tracing
Yazhuo Zhang, Rebecca Isaacs, Yao Yue, Juncheng Yang, Lei Zhang, Ymir Vigfusson
The 14th ACM Symposium on Cloud Computing (SoCC), 2023
[code]

[SOSP] FIFO queues are all you need for cache eviction
Juncheng Yang, Yazhuo Zhang, Ziyue Qiu, Yao Yue, K. V. Rashmi.
The 29th ACM Symposium on Operating Systems Principles (SOSP), 2023
[website][blog][slides][video]

[HotOS] FIFO Can be Better than LRU: the Power of Lazy Promotion and Quick Demotion
Juncheng Yang, Ziyue Qiu, Yazhuo Zhang, Yao Yue, K. V. Rashmi.
The 19th Workshop on Hot Topics in Operating Systems (HotOS), 2023
[slides]

Services

ACM Transactions on Storage (TOS) 2025: Reviewer
SESMAE'25: Program Committee
ICDCS'25: External Reviewer
ATC'24: Artifact Evaluation Committee
OSDI'24: Artifact Evaluation Committee
Eurosys'24: External reviewer
ATC'21: External reviewer
HotCloud'20: External reviewer
LADIS'21: Volunteer

Teaching

2025 Spring Cloud Computing Architecture
2021 Spring CS326 Analysis of Algorithms
2020 Fall CS551 System programming
2020 Spring CS170 Introduction to Computer Science
2018 Fundamentals of Computers
2017 Database
2017 Parallel Algorithms and Programming

Life

Outside research, I enjoy playing piano, tennis, swimming, running, exploring new places, and interior design.