Jiaxin Shan
Software Engineer at ByteDance
Seattle, Washington, United States
Summary
🤩
Rockstar🎓
Top SchoolJiaxin Shan is a seasoned software engineer with 12 years' experience building large-scale distributed systems, currently driving resource management, serverless inference, and GPU cluster utilization at ByteDance. Previously at AWS EKS and eBay, Jiaxin led ML and big-data initiatives on Kubernetes, helping dozens of teams productionize ML platforms and integrating cloud-native components like EKS, Istio, FSx, SageMaker, EMR and Athena. He combines deep systems expertise (resource managers, RPC, schedulers) with hands-on MLOps and DevOps work—contributing significantly to high-profile open-source projects such as Kubeflow, KubeRay, Kubernetes autoscaler and vLLM. Notably, he has implemented CI/CD automations, AWS-specific integrations, and dynamic LoRA support for LLM serving, reflecting both backend infrastructure depth and ML-serving pragmatism. Based in Seattle, Jiaxin blends research-grade thinking (papers on resource management and CXL memory) with pragmatic engineering that optimizes cost and performance at cloud scale.
12 years of coding experience
7 years of employment as a software developer
Master of Science (M.S.) Computer and Information Sciences General, Master of Science (M.S.) Computer and Information Sciences General at University of PittsburghBachelor’s Degree Computer and Information Sciences, Bachelor’s Degree Computer and Information Sciences at JiangNan University
Github Skills (54)
ec10
autoscaling10
amazon-eks10
pytorch10
kubernetes10
docker10
controller10
python10
apidoc10
testing10
kustomize10
huggingface-hub10
kubeflow10
machine-learning10
cloud-infrastructure10
Programming languages (24)
Github contributions (5)
ray-project/kuberay
Aug 2021 - Dec 2022
A toolkit to run Ray applications on Kubernetes
Role in this project:
DevOps Engineer & Backend Developer
DevOps Engineer & Backend DeveloperContributions:5 releases, 429 reviews, 73 commits in 1 year 4 months
Contributions summary:Jiaxin's commits primarily focus on setting up and configuring the Ray operator project within the Kubernetes environment. They initialized the project using `kubebuilder`, defined the RayCluster API, and modified the code structure to be project-friendly. Furthermore, the user implemented the core controller logic for RayCluster, including service and pod reconciliation, and addressed status update issues to ensure correct cluster management. The user also made key improvements to the codebase, including adding support for in-tree autoscaling and fixing flaky test issues.
raydeep-learningapachemachine-learningkubernetes
kubeflow/trainer
Nov 2018 - Nov 2021
Distributed ML Training and Fine-Tuning on Kubernetes
Role in this project:Backend & DevOps Engineer
Backend & DevOps EngineerContributions:8 releases, 169 reviews, 37 commits in 3 years
Contributions summary:Jiaxin primarily contributed to the unification of code structure within the training job API, enhancing the project's maintainability and consistency. They added documentation and registered the API within the job framework. Furthermore, the user updated the codegen scripts to generate defaulters and openAPI specifications for various frameworks. These changes involved modifications to code generation, API definitions, and controller logic, indicating work across backend and DevOps aspects of the project.
xgboostkubernetesmachine-learningtrainingai
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.