معرفی
Andy Pavlo is an Associate Professor with Indefinite Tenure in the Computer Science Department at Carnegie Mellon University's School of Computer Science. He is an active member of the CMU Database Group and the Parallel Data Laboratory, where he leads research in database management systems with a focus on self-driving architectures, transaction processing, and large-scale analytics. His work bridges academic research and industry applications through projects like NoisePage, OtterTune (which he co-founded and served as CEO before it ceased operations), and Peloton.
Dr. Pavlo's research interests span database management systems with particular emphasis on autonomous database architectures that can self-tune and optimize without human intervention. His work explores transaction processing systems that can handle high-throughput workloads while maintaining consistency, and large-scale data analytics techniques that efficiently process massive datasets. He has made significant contributions to query optimization, database extensibility, and automatic database tuning using machine learning techniques. His recent work on database extensibility revealed critical issues in PostgreSQL's extension ecosystem, showing that approximately 16% of extensions are incompatible with at least one other extension due to API violations and memory errors.
His research output demonstrates a consistent focus on practical database systems challenges, with recent publications examining database extensibility, user-defined function optimization, and the cyclical nature of database research. The articles show a strong trend toward making database systems more autonomous, with increasing integration of machine learning techniques for automatic tuning and optimization. His work often combines deep theoretical analysis with practical implementation in open-source systems.
- Dijkstra Award 2024 for contributions to database systems research
Dr. Pavlo actively mentors graduate students, with current advisees including Wan Shen Lim, William Zhang, and Sam Arch (co-advised with Todd Mowry). His former students have gone on to successful careers in both industry and academia. He has secured significant research funding through CMU's affiliate program with major database companies including ClickHouse, DataStax, dbt, Firebolt, MotherDuck, RelationalAI, SingleStore, Spiral, PingCAP/TiDB, Yellowbrick, and Yugabyte. His research is supported by these industry partnerships and likely includes NSF funding given his active participation in the database research community.
At CMU, Dr. Pavlo leads the Database Group and organizes several seminar series including "SQL or Death," "Database Building Blocks," and "ML⇄DB Technical Talks." These seminars bring together researchers and practitioners to discuss cutting-edge developments in database systems. He also runs a summer research internship program that has attracted students for multiple consecutive years, indicating a strong research group with ongoing projects and funding.
