Summary
Zhengyuan Zhu is a machine learning engineer with a decade of experience building production-grade AI systems, currently designing the multi-agent orchestration behind Adobe Experience Platform’s enterprise assistant. He combines a strong NLP research background—15 papers on misinformation detection, fact-checking, and knowledge graphs during a PhD at UT Arlington—with hands-on systems engineering that routes and composes dozens of specialized agents in real time for tens of thousands of users. His MuRAR multimodal retrieval-augmented reasoning work won Best Demonstration at COLING 2025 and led to an Adobe patent, reflecting a knack for turning research into scalable product features. Pragmatic and detail-oriented, he improved agent routing accuracy to 89% through context-aware ranking and builds grounded response synthesis pipelines that prioritize trust and provenance. Based in the Bay Area, he balances academic rigor with production impact, focused on making multi-agent AI reliable at enterprise scale.
11 years of coding experience
5 years of employment as a software developer
Doctor of Philosophy - PhD, Computer Science, Doctor of Philosophy - PhD, Computer Science at The University of Texas at Arlington
Bachelor's Degree, Information Management and Information Systems, Bachelor's Degree, Information Management and Information Systems at Tianjin University of Finance and Economics