Pei Ran Li

About Pei Ran Li

9+ years of experience with design, implementation, and tests of scalable RESTful APIs and infrastructures for cloud - SaaS/PaaS- 9+ years in Java, Python, Typescript, Spring Framework, AWS, MySQL, Hibernate, Cassandra- Delivered fault tolerant and highly available cloud software applications through CI/CD automations, rollbacks, monitoring and alerting- Excellent communication and documentation skills

Experience

1. Software Engineer

   Amazon

   May 2020 — Present · Seattle, WA, US

   Built and maintained the inference infrastructure for Amazon\'s conversational AI customer service chatbots, leveraging large language models (LLMs)- Developed and owned the deployment pipelines for integrating and deploying LLM models into production environments, including rigorous testing and validation stages- Designed and implemented real-time inference services for large-scale conversational AI systems, with a focus on optimizing model performance for low latency and high transactions per second (TPS) in a customer service context- Designed and implemented the ML inference service that orchestrates the interaction between multiple models, such as issue prediction, item recommendation, and response generation, to provide coherent and contextual responses- Integrated with various cross-team services to enable retrieval-augmented generation (RAG) systems, combining language models with external knowledge sources and actuation capabilities to deliver informational and action-oriented responses- Enhanced observability and logging mechanisms to proactively identify and troubleshoot issues, and maintain dialogue state for offline training and model improvement- Collaborated closely with Product Managers, UX designers, Applied Scientists, and experienced Software Development Engineers to effectively apply machine learning models and deliver high-quality conversational AI experiences- Designed and implemented a scalable RPC service in Java and Spring to translate languages in real time- Designed and executed a plan to transition the single tenant monolithic application to microservices with multi-tenant support and data isolation and high availability- Created a platform with Docker containers, managed by Kubernetes, that can be onboarded by ML scientists to train and deploy large language models- Implemented and deployed reliable infrastructures and AWS resources using CDK with CI/CD pipelines and automatic rollbacks

Education

• University of Waterloo

  Bachelor’s Degree, Computer Engineering

  2011 — 2016

Skills

• Jquery
• Web Applications
• Windows Scripting
• Jdbc
• J2ee Application Development
• Javabeans
• Test Planning
• Kendo
• Agile Methodologies
• Angularjs
• Test Cases
• User Acceptance Testing
• Xhtml
• Test Driven Development
• Javascript
• C++
• Agile & Waterfall Methodologies
• Sql
• Software Development
• Scrum
• Junit
• Css
• Sprint Planning
• Html
• Visual C#
• Visual Studio
• C#
• Php
• Java
• Microsoft Sql Server
• Git
• Entity Framework
• Jsp