Beibin Li
Member Of Technical Staff at xAI
Seattle, Washington, United States
Summary
🤩
Rockstar🎓
Top SchoolBeibin Li is a lead scientist and engineering leader with 11 years of experience building and shipping large-scale ML and agentic AI systems across industry labs including xAI, Meta, and Microsoft Research. He has led coding and post-training efforts—driving RL, evaluation, and training pipelines for Grok—and now heads self-evolution and coding teams at MiroMind.ai. His work blends deep research (PhD, UW) with hands-on backend development, including contributions to Microsoft’s popular autogen framework improving OpenAI API resilience and code extraction. Comfortable moving models from lab to product, he focuses on robustness, tooling, and developer-facing systems that make advanced reasoning and coding models reliable in practice. Based in Seattle, he pairs academic rigor with pragmatic engineering and a knack for surfacing subtle edge cases in model integrations.
11 years of coding experience
5 years of employment as a software developer
Doctor of Philosophy - PhD, Computer Science and Engineering, Doctor of Philosophy - PhD, Computer Science and Engineering at University of WashingtonBachelor of Science (B.S.), Computer Science, Mathematics, Bachelor of Science (B.S.), Computer Science, Mathematics at University of MichiganMathematics, Mathematics at Rhodes CollegeShenzhen Middle School
Github Skills (7)
error-handling10
autogen10
openai-api10
agent-framework10
python10
regex9
regular-expression9
Programming languages (6)
Github contributions (5)
microsoft/autogen
Sep 2023 - Nov 2024
A programming framework for agentic AI 🤖 PyPi: autogen-agentchat Discord: https://aka.ms/autogen-discord Office Hour: https://aka.ms/autogen-officehour
Role in this project:
Backend Developer
Backend DeveloperContributions:1 release, 177 reviews, 52 PRs in 1 year 1 month
Contributions summary:Beibin's commits focus on enhancing the `autogen` framework, primarily involving the backend logic and error handling related to the OpenAI API calls. Specifically, the user added error handling for authentication issues when using different OpenAI configurations, improving the framework's resilience. Additionally, the user contributed to code extraction functionality, enabling the detection of single-line code snippets and refining code parsing logic. This work suggests a focus on improving the robustness and functionality of the agent framework.
agenticagentic-agiagentsaiautogen
microsoft/OptiGuide
Jul 2023 - Feb 2025
Large Language Models for Supply Chain Optimization
Contributions:4 reviews, 21 PRs, 51 pushes in 1 year 6 months
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.