This document summarizes a presentation about the Boss database, a petascale database for neuroscience powered by serverless analytics on AWS. It discusses how serverless computing allows running code without managing servers, paying only for compute time used. It then describes the JHU APL/IARPA MiCRONS project which uses AWS serverless services like Lambda, Step Functions, DynamoDB and others to build a database for large-scale neuroscience research that is scalable, flexible and easy to manage.