Xun Zhou
Quantitative Developer
Shanghai, Shanghai, China
Summary
🤩
Rockstar🎓
Top SchoolRick Zhou is a Quantitative Developer with nine years of experience building ML systems, distributed service infrastructure, and compiler-accelerated model deployments. He blends research and production chops from CMU’s Catalyst group with hands-on engineering at LinkedIn, where he helped integrate gRPC, service discovery, and next-gen RPC traffic management using Envoy. His open-source contributions span high-impact projects—rest.li consistent hashing and balancer work at LinkedIn, deep improvements to TVM’s compiler stack, and practical enhancements to MLC-LLM and Web-LLM that enable native and in-browser LLM inference. Now based in Shanghai and working at Minghong Investment while continuing CMU research, he moves fluidly between low-level C++/compiler fixes and large-scale distributed design. Colleagues rely on him for subtle performance optimizations and robustness fixes that often live at system boundaries (e.g., KV cache, retries, and balancer strategies).
10 years of coding experience
4 years of employment as a software developer
Master of Science - MS Computer Science, Master of Science - MS Computer Science at Carnegie Mellon University
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (34)
algorithm10
consistent-hashing10
tvm10
algorithms10
javascript10
c-language10
webworker10
language-model10
python10
command-line-interface10
machine-learning10
load-balancing10
data-structure10
java10
llm10
Programming languages (10)
Github contributions (5)
linkedin/rest.li
Jul 2019 - Nov 2022
Rest.li is a REST+JSON framework for building robust, scalable service architectures using dynamic discovery and simple asynchronous APIs.
Role in this project:
Back-end Developer
Back-end DeveloperContributions:270 reviews, 52 commits, 91 PRs in 3 years 3 months
Contributions summary:Xun primarily contributed to the implementation of a bounded-load consistent hashing algorithm and a consistent hash ring simulator. They added functionality for the D2 balancer, including monitoring metrics for relative strategy in DegraderLoadBalancerStrategyV3Jmx and fixing retry client bugs. Further commits indicate enhancements to client-side retry logic and configuration, as well as cluster subsetting functionality.
asynchronous
mlc-ai/mlc-llm
Oct 2023 - Oct 2024
Universal LLM Deployment Engine with ML Compilation
Role in this project:Back-end Developer & DevOps Engineer
Back-end Developer & DevOps EngineerContributions:14 reviews, 53 PRs, 22 pushes in 1 year
Contributions summary:Xun primarily focused on enhancing the C++ CLI for the MLC-LLM project, adding features like the model library path override and improving error messages. Their work involved modifying the `cli_main.cc` file to incorporate the new functionality and updating documentation. Additionally, the user made modifications to the Python chat module, suggesting a contribution to the overall deployment and execution of the LLM.
llmmachine-learningmachine-learning-compilationlanguage-modeltvm
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.