avatar

Yuangang Li

Mater Student in Computer Science at the University of Southern California. & Interests include Cloud Computing, Machine Learning, and Distributed Systems.

Work Experience

Software Developer | USC iLab Laboratory Los Angeles, CA Mar.2023 – Aug.2023
Python, MLOps, Docker, Github Action, Pytorch, Pytest
  • Contributed to “Ablator,” an open-source Deep Learning framework used by 30+ USC researchers for horizontal scaling of ablation experiments and hyperparameter tuning, encompassing 70 pull requests. Ablator
  • CreatedCI/CD pipeline with GitHub Actions, Implemented and documented tests with Pytest, achieving 95%+ coverage.
  • Developed and released a Python API package for RClone on PyPI, streamlining cross-platform usage without pre-installation and featuring automatic binary selection. python-rclone

Infrastructure Engineer | SenseTime (Top Tier Public AI Company)Jan.2022 – Dec.2022
Kubernetes, Docker, Go, CI/CD, CRD, Operator-SDK, Helm3, Prometheus, Grafana
  • Developed “RocketMQ as a Service,” akin to “RabbitMQ as a Service” in AWS Marketplace, offering fully managed RocketMQ clusters in SaaS Platform, resulting in a 100% increase in RocketMQ service creation speed.
  • Utilized Operator SDK to construct a Kubernetes-based RocketMQ Operator and Custom Resource Definitions (CRD), automating lifecycle management and saving 10+ hours in manual operations per week.
  • Employed Helm3 to package components into Helm charts, simplifying Kubernetes deployment and improving resource management for the RocketMQ Operator project.
  • Implemented Prometheus and Grafana for real-time monitoring of critical service metrics and node health.
  • Automated workflows including unit tests, image builds, and Helm3 Chart updates via GitLab CI/CD.

Software Engineer | Institute of Software, Chinese Academy of Sciences (Top Research Institute)May.2021– Aug.2021
Docker, Node.js, JSON, Vue.js, RobotFramework
  • Participated in the R&D of a web-based automated testing tool employing Record and Playback technology. This tool allowed the recording, editing, execution, analysis, and result generation of user actions within test cases.
  • Used Node.js, transformed user actions recorded in JSON into Robot Framework scripts, achieving a 95% operation conversion coverage rate, to facilitate Playback testing.
  • Containerized the program with Docker and automated DevOps pipeline for development efficiency improvement.
  • Employed Vue.js and AceEditor to fabricate custom components and layouts, enhancing front-end usability.
  • The implementation of the tool saves 15+ hours per week and improves end-to-end testing efficiency by 300%.

Software Developer | Xiaoniu Translations (Beijing) Technology Co., Ltd.Jan.2021– Apr.2021
Java, SpringBoot, Spring, Java Persistence API, Maven, Nginx, MySQL, Git, Java Application
  • Developed an AI document translation system with Java/Spring/Maven, independently created a PDF/XML module garnering 30,000 MAUs, and optimized performance using Nginx and Git.