# Xiang Li > Software Engineer at Coupang | Ex-Retool | Ex-Googler | Venture Scout Location: San Francisco Bay Area, United States Profile: https://flows.cv/xiangli I build products at the intersection of AI, design, and engineering—where great taste matters as much as great models. With rapid progress in LLMs and multimodal AI, I believe the next generation of products won’t just need strong tech—they’ll need the kind of intuition, clarity, and user empathy that comes from thoughtful design. Right now at Coupang, I'm working on reimagining search with LLMs and image-based ML. We're building smarter, more intuitive search experiences that understand both language and visuals—making it easier for millions of users to find what they love. Before this, I was at Retool, where I led a major overhaul of our source control infrastructure—improving scalability, cutting stability issues by 80%, and speeding up deployments by 91%. Earlier at Google, I led the Discover content understanding team, applying models like BERT and MUM to deliver fresher, more relevant recommendations. I'm especially excited about what’s next: AI-native experiences, multimodal interfaces, and the tools that will shape how we build in an AGI world. 📩 bylixiang+jobs@gmail.com ## Work Experience ### Software Engineer @ Coupang Jan 2024 – Present | Mountain View, California, United States • Led the development of an AI-powered search experience at Coupang, enhancing product discovery through multimodal approaches. • Utilized Triton, vllm, TensorRT, and ONNX for efficient LLM model serving, optimizing performance. • Implemented generative image editing techniques to directly impact e-commerce, improving user engagement. ### VC Scout @ UpHonest Capital Jan 2023 – Present ### Software Engineer @ Retool Jan 2022 – Jan 2024 Engineering Leader on the enterprise features of developer life cycle including building, collaborating, testing, releasing and monitoring. - Led the broadening of Retool's source control integration, incorporating all key management providers: GitHub, GitLab, Bitbucket, Azure Repos, and AWS CodeCommit. This unlocked many business opportunities by enhancing platform versatility and compatibility. - Refined the Retool platform by embedding source control functionalities, ensuring extensive support across the platform's elements, creating an developer friendly environment. - Created an innovative customer app replay flow, significantly improving the customer support journey by enabling more efficient and effective resolution of user inquiries and issues. Saving hundreds of hours of support time and improved root cause identification from multiple days to minutes. ### Senior Software Engineer | Tech Lead @ Google Jan 2018 – Jan 2022 | Mountain View, California, United States As a Tech Lead at "Google Discover" https://support.google.com/webmasters/answer/9046777?hl=en. Train machine learning models with large language models like BERT, MUM to improve recommendation system performance. - Analyzing product business needs and developing scalable solutions - Designing and implementing machine learning pipelines, and training models to improve user experience and drive product growth - Working cross-functionally with data scientists and program managers to understand and deploy machine learning models - Leading project technical roadmaps, balancing short and long-term technology investments, and delegating tasks to team members - Mentoring and training team members on design techniques, modeling standards, and project directions ### Software Engineer @ Google Jan 2016 – Jan 2018 | San Francisco Bay Area Full stack engineer at "Google My Business" https://www.google.com/business/ - Design frontend/backend systems and components with performance in mind that enable positive user experiences and measurable business growth - Design RESTful APIs that's part of https://developers.google.com/my-business/reference/rest and complies with https://google.aip.dev/ standard - Build and use tools and instrumentation to help monitor and analyze performance - Coordinate with product managers in all part of the product life cycle ### Software Engineer @ Silvus Technologies, Inc. Jan 2014 – Jan 2016 Fullstack Engingeer - Front-End: Use Javascript framework like ReactJS for user interface, implement jsonrpc client/server, and design responsive web app with bootstrap - Back-End: Write python,c and shell scripts code for backend api, implement FastCGI with lighttpd - Automate front-end development with Gulp.js, optimize code for max speed and scalability - Reduced web page rendering time by 80%, Reduced memory and CPU usage by removing uncollected objects ### Software Engineer Intern @ Silvus Technologies, Inc. Jan 2013 – Jan 2013 - Implement realtime position map features with open resources(Openstreet Map, Google Map, Openlayers), created topology graph with jit.js - Built Auto web test framework for web interface with Selenium(Selenium webdriver) ## Education ### Stanford LEAD Professional Certificate Stanford University Graduate School of Business ### Master of Science - MS in Electrical and Electronics Engineering UCLA ### Bachelor of Science (BS) in optical engineering Zhejiang University ## Contact & Social - LinkedIn: https://linkedin.com/in/xiang-li-zju --- Source: https://flows.cv/xiangli JSON Resume: https://flows.cv/xiangli/resume.json Last updated: 2026-03-22