معرفی
Fuheng Zhao is a newly appointed Tenure-Track Assistant Professor at the University of Utah's School of Computing, transitioning from his recently completed Ph.D. in Computer Science at the University of California, Santa Barbara. His research bridges database systems and artificial intelligence, focusing on novel approaches to data management at scale.
Dr. Zhao's educational background includes a Ph.D. from UC Santa Barbara where he was co-advised by Professors Divyakant Agrawal and Amr El Abbadi. His dissertation work centered on developing algorithms and systems for efficient big data management, with a particular focus on integrating large language models into traditional database systems.
Research Focus: Dr. Zhao's work explores the intersection of database systems and AI, with three primary thrusts: (1) Natural language interfaces to databases including robust NL2SQL systems and hybrid query processing that combines relational operators with LLM calls; (2) High-performance streaming algorithms and sketches for real-time analytics with accuracy guarantees; (3) Advanced indexing structures including B-trees, LSM-trees, and vector indexes for embedding-based data. His approach emphasizes both theoretical grounding and systematic engineering, aiming to build systems that balance token efficiency, latency, and accuracy.
His publication record shows a clear evolution from foundational streaming algorithms and sketching techniques toward increasingly sophisticated integration of LLMs with database systems. Recent work demonstrates innovative approaches to resolving ambiguities in natural language to SQL translation, developing cost models for token-accuracy trade-offs, and exploring how databases should evolve for AI agent users.
- Microsoft Research PhD Fellowship
- Charles Dana Fellowship
- Research gifts from Snowflake
Dr. Zhao is actively recruiting self-motivated students interested in building the next generation of intelligent data systems. He emphasizes passion for research as the most important quality, with preferred background in computer science, particularly algorithms, data structures, systems, and databases. His lab focuses on concrete problems including query optimization for LLM operators, vector-relational fusion, and designing databases for AI agents. He has made significant contributions to the FoundationDB open-source project during his internships at Snowflake, focusing on monitoring, caching, and scheduling disk I/O.
