Yunpeng Li

• Lead DevOps and infrastructure initiatives for Apple News, Weather, and Stocks, supporting large-scale production services
• Designed and implemented internal tooling, including AI/LLM-powered systems and chat-based interfaces to improve developer productivity and reduce operational friction
• Built and integrated retrieval-based systems (RAG) and embedding pipelines for intelligent query and automation workflows
• Developed agentic workflows enabling systems to execute infrastructure actions and respond to natural language queries
• Built and operated scalable deployment infrastructure using Kubernetes and AWS services
• Automated CI/CD pipelines and operational workflows, improving reliability and reducing manual intervention
• Enhanced system observability, deployment efficiency, and operational visibility across services
• Owned end-to-end DevOps responsibilities, including infrastructure provisioning, deployment, monitoring, and tooling
• Provided technical leadership and guidance in system design, platform architecture, and AI-enabled workflow integration

• Designed and built centralized logging system for CI pipelines, improving debugging, observability, and build analytics
• Built next-generation CI infrastructure on AWS, improving efficiency and reducing cost
• Provided senior-level technical guidance and solutions across teams

• Designed and implemented Apple News automation pipeline, reducing execution time from hours to minutes
• Built scalable infrastructure using Kubernetes, Jenkins, and VM Cloud, reducing qualification time from daily to hourly
• Created Kubernetes-based development helper service, significantly improving developer productivity
• Built auto-scaling UI testing platform using Selenium and Dockerized headless browsers
• Improved system stability and reduced execution time of testing workflows
• Enabled PR-based concurrent testing, increasing developer efficiency by over 100%
• Contributed to hiring and mentoring engineers

• Built cloud datacenter infrastructure and services
• Designed VM and container image catalog systems with high performance and high availability
• Developed provisioning systems using KVM and QEMU
• Contributed to OpenStack

• Modernized QA automation pipelines with on-demand provisioning, significantly improving resource utilization
• Built cloud-based test automation pipeline with fault tolerance and CI integration
• Developed SDKs and internal tools used across teams
• Built infrastructure monitoring and recovery systems using Node.js and NoSQL
• Improved system stability and reduced flaky tests
• Mentored interns and led successful project delivery

• Built high-performance web testing systems using headless browsers (PhantomJS)
• Developed frontend automation testing using JavaScript, jQuery, and Node.js
• Improved development workflows and testing efficiency
• Built internal tools across multiple programming languages

AI-Powered DevOps Platform

Designed and implemented a chat-based system that replaces traditional DevOps portals with a unified interface Built LLM-powered assistant capable of executing infrastructure actions and answering system-related queries Implemented embedding-based retrieval (RAG) pipelines for contextual and accurate responses Developed agentic workflows to automate multi-step DevOps operations Integrated local and remote LLMs (e.g., DeepSeek, Ollama) into production workflows Significantly improved developer productivity and reduced time spent navigating tools Simplified complex infrastructure workflows into a natural language interface

Bachelor of Engineering (B.Eng.), Engineering Systems & Computing (Co-op, Honours)

University of Guelph, Canada · June 11, 2013