Shengbo Song
Software Engineer at Tencent
Haidian District, Beijing, China
Summary
👤
Senior🎓
Top SchoolShengbo Song is a seasoned software engineer with 11 years of experience building infrastructure and backend systems, currently at Tencent in Beijing. He has deep hands-on expertise in GPU scheduling, container runtimes and Kubernetes integration—having implemented GPU scheduling for the Gaia system and designed an API server to translate internal APIs to the Kubernetes API. Earlier work spans search-engine components on Hadoop/HBase, content deduplication and ML-based content classification, showing strong systems and data-processing skills. His open-source contributions to tkestack projects improved GPU management and CUDA/container support, including refactoring for broader CRI compatibility and modernizing build processes. Shengbo moves quickly from concept to production, enjoys challenging problems, and often focuses on making complex platform features easier for end users and operators. He combines low-level systems knowledge with practical DevOps experience, including custom CI plugins and automation that sped release workflows.
11 years of coding experience
5 years of employment as a software developer
Bachelor’s Degree, Computer Science, Bachelor’s Degree, Computer Science at Harbin Institute of Technology
English, Chinese
Github Skills (14)
cuda10
kubernetes10
docker10
go10
containerization10
dockers10
kubernetes-pods10
devops9
c119
c179
cri-o8
linux8
container-runtime-interface8
systemd7
Programming languages (9)
Github contributions (5)
tkestack/vcuda-controller
Nov 2019 - Nov 2022
Role in this project:
Back-end & DevOps Engineer
Back-end & DevOps EngineerContributions:1 release, 2 reviews, 31 commits in 3 years
Contributions summary:Shengbo primarily focused on enhancing the project's CUDA and containerization capabilities. They fixed issues related to container runtime support and updated the codebase to accommodate different container environments. Furthermore, the user integrated host network support for image building and addressed bugs associated with CUDA initialization. They also incorporated the latest CUDA 11 and 11.5.1 functionality and resolved a recursive problem within the project.
tkestack/gpu-manager
Nov 2019 - Dec 2021
Role in this project:Back-end & DevOps Engineer
Back-end & DevOps EngineerContributions:3 releases, 69 commits, 19 PRs in 2 years 2 months
Contributions summary:Shengbo primarily focused on refactoring and improving the `gpu-manager` codebase, switching from `glog` to `klog` and removing dependencies on Kubernetes code. They also implemented support for the CRI interface, indicating an effort to broaden the compatibility of the system. The user contributed to build processes, including setting up the build environment with host networking, and upgrading dependencies.
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.