Self-Driving Database Management Systems

Published in Conference on Innovative Data Systems Research, 2020

Summary: We explored constructing a self-driving database management system: Peloton DB

This excerpt from the abstract does a fantastic job of summarizing our work:

This is different than earlier attempts because all aspects of the system are controlled by an integrated planning component that not only optimizes the system for the current workload, but also predicts future workload trends so that the system can prepare itself accordingly. With this, the DBMS can support all of the previous tuning techniques without requiring a human to determine the right way and proper time to deploy them. It also enables new optimizations that are important for modern high-performance DBMSs, but which are not possible today because the complexity of managing these systems has surpassed the abilities of human experts.

This was part of my undergraduate research, where I analyzed real-world data queries, anonymized data, and experimented with featurizations of the query. The output of my work was fed to a downstream pipeline.