Recent work on "learned indexes" has changed the way we look at the decades-old field of DBMS indexing. The key idea is that indexes can be thought of as "models" that predict the position of a key in a dataset. Indexes can, thus, be learned. The original work by Kraska et al. shows that a learned index beats a B+Tree by a factor of up to three in search time and by an order of magnitude in memory footprint. However, it is limited to static, read-only workloads. In this talk, I will talk about a new learned index called ALEX which addresses practical issues that arise when implementing learned indexes for workloads that contain a mix of point lookups, short range queries, inserts, updates, and deletes. ALEX effectively combines the core insights from learned indexes with proven storage and indexing techniques to achieve high performance and low memory footprint. ALEX presents a key step towards making learned indexes practical for a broader class of database workloads with dynamic updates.

Dr. Jaeyoung Do is a Senior Researcher at Microsoft Research, where he is working on building large-scale systems with ML-enhanced data structures and algorithms, and improving such systems with software and hardware co-optimization for big data and AI processing in the context of cloud computing. He received a Ph.D. in 2012 and an M.S. in 2009 from the University of Wisconsin-Madison, USA, and a B.S. in 2007 from Korea Advanced Institute of Science and Technology (KAIST), Korea, all in Computer Sciences. During his Ph.D. study, he explored several advanced architectures to effectively integrate flash SSDs into existing database management systems. His work on using flash SSDs to extend a main-memory DBMS buffer pool has been shipped to Microsoft SQL Server 2014.