Kevin Xiang Li

• LinkedIn
• GitHub
• Resume

Current: Stanford MSCS, Prev: UMich CS, Intern at Nvidia

I'm interested in post-training, particularly synthetic data generation for coding agents.

In my past lives, I've also deployed a mobile app with 30K+ installs across the App Store, Play Store, and AppGallery; built AR/VR apps with Unity/Unreal for Meta Quests; and published at top programming language conferences.

Projects

• SWE-smith Multilingual: Scaling Multilingual Data for Software Engineering Agents

  

• Async RL from Scratch on TPUs

  

• Accelerating Vision-Language Models in SGLang

  

• VideoMultiAgents: A Multi-Agent Framework for Video QA

  

• Live-It: Image-to-3D Scene Generation

• Static Retrieval

  

• Words.hk app

  

• Arborist program synthesizer

  

• A2GO in AR

  

• DIPS: Efficient Multi-Criteria Cantonese Word Segmentation

  

• CSE Simulator in VR

  

• Hangeul Korean input method

  

• RenovAR

  

• Wukong Kinematics

• Pinccer Physics Simulation