Summary
Linfeng Song is a Staff Engineer at Google with a decade of experience building agent frameworks, context management, and scalable RL systems for high-stakes reasoning tasks. He led RLHF and PPO work on large LLMs at Tencent, created a self-evolving search-and-reasoning pipeline adopted by product teams, and helped push automated theorem proving forward with an agentic system that solved post-2000 IMO problems using modest compute. His background blends deep academic rigor—a PhD in Computer Science—with hands-on research and production experience across ML, NLP, and reinforcement learning. Based in Bellevue, Washington, he brings a rare combination of frontend sensibilities and research-first engineering that enables both robust infrastructure and advanced model-driven products.
10 years of coding experience