Xinyuan Huang
Technical Leader at Cisco Meraki
San Jose, California, United States
Summary
👤
Senior🎓
Top SchoolXinyuan Huang is a technical leader with 12 years of experience building AI/ML systems for edge and cloud, currently driving camera intelligence platforms at Cisco Meraki and broader cloud AI research at Cisco. He chairs an MLCommons working group and was an early contributor to MLPerf, bringing benchmark-grade rigor to model training and compliance logging. His open-source contributions span high-profile projects like MLCommons' training benchmarks and Kubeflow’s Kubebench, where he improved deployment prototypes, parameterization, and verification tooling. Xinyuan blends hands-on ML engineering—model/data fixes and beam-search tweaks—with MLOps expertise in Kubernetes-centric benchmarking and deployment. Trained in electronic engineering and machine learning across Fudan, Birmingham, and UCL (Distinction), he pairs deep academic grounding with practical production delivery. Colleagues rely on him to translate research-grade models into auditable, scalable systems that run at the edge.
12 years of coding experience
Bachelor of Engineering (BEng), Electronic Engineering, 1st Class Honours, Bachelor of Engineering (BEng), Electronic Engineering, 1st Class Honours at The University of BirminghamBachelor of Science (BSc), Electronic and Information Science and Technology, Bachelor of Science (BSc), Electronic and Information Science and Technology at Fudan UniversityMaster of Science (MSc), Machine Learning, Distinction Award, Master of Science (MSc), Machine Learning, Distinction Award at University College London, U. of London
English, Chinese
Github Skills (20)
kubernetes10
benchmark10
python10
apidoc10
kubeflow10
machine-learning10
benchmarking10
mlp10
mlops10
api10
tensorflow10
perf10
jsonnet10
kubernetes-pods10
transformer10
Programming languages (10)
Github contributions (5)
mlcommons/training
Oct 2018 - Nov 2019
Reference implementations of MLPerf™ training benchmarks
Role in this project:
ML Engineer
ML EngineerContributions:8 commits, 12 PRs, 25 comments in 1 year 1 month
Contributions summary:Xinyuan contributed to the MLPerf training benchmark implementations within the `mlcommons/training` repository. Their work focused on adding compliance logging for the Transformer model, modifying scripts to install necessary packages, and fixing related issues. Furthermore, they updated processed data filenames and made changes to the model's beam search functionality. They also added a submission helper to assist with verification and extraction.
implementationsbenchmarkingdeep-learningmlperfmachine-learning
kubeflow/kubeflow
Apr 2018 - Apr 2019
Machine Learning Toolkit for Kubernetes
Role in this project:MLOps Engineer
MLOps EngineerContributions:10 commits, 12 PRs, 43 comments in 11 months
Contributions summary:Xinyuan primarily contributed to the Kubebench project, which is a machine-learning benchmark tool for Kubernetes. Their work involved adding and modifying configurations for the Kubebench dashboard, including deployments and service definitions using Jsonnet. They updated the Kubebench job prototype with changes to default parameter values, and introduced a new package and associated code, demonstrating an understanding of the project's architecture and deployment process. They also addressed the parameters used by the prototypes, fixing hardcoded variables and adding the ability to accept parameters.
pythondata-sciencenotebookmachine-learningmlops
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.