NIPS 2018 Expo Demo

Dec. 2, 2018

Expo Schedule »

Amazon’s internal hosted notebook service accelerating ML adoption, collaboration, and education

Sponsor: Amazon

Abstract:

With Jupyter and Zeppelin frameworks, we have seen notebooks gain wide adoption in the data science and ML community. At Amazon, we leverage notebook technologies to bring the ability to learn, experiment and productionalize ML applications to everyone without requiring users to become technical experts to make data-driven decisions. We will demonstrate Eider: Amazon’s internal on-demand notebook service for data science that accelerates data-driven decision making and machine learning use across the company.

We will showcase key features of the platform that enable sharing and collaboration of ML solutions, instant access to compute ready with ML package/libraries pre-installed, easy access to AWS services, instant reproducibility and automation of notebook solutions. We will walk attendees through examples of how Eider is used to accelerate ML at each stage of the ML life-cycle. Specifically we will showcase features that accelerate the use case of ML education programs at Amazon.