Summary
Xingchen Wan is a Senior Research Scientist at Google DeepMind with a decade of experience pushing large language models toward practical, secure, and agentic applications. His work spans coding agents, post-training techniques for Gemini, and enterprise-focused LLM improvements developed at Google Cloud—projects that include Vertex AI Prompt Optimizer and multiple NeurIPS/ACL/ICLR papers. He combines rigorous academic training (MEng and PhD from Oxford) with hands-on production impact, earning a Google Cloud Tech Impact Award for prompt optimization. Notably, he has led research on automating multi-agent LLM systems and agentic RAG, bridging cutting-edge research with deployable cloud products. Based in San Francisco, he brings research depth and applied engineering to scale LLMs for real-world and security-sensitive use cases.
10 years of coding experience
2 years of employment as a software developer
PhD Machine Learning, PhD Machine Learning at University of Oxford
Singapore-Cambridge GCE A-Levels, Singapore-Cambridge GCE A-Levels at River Valley High School