AWS Clean Rooms Lab

This lab will walk you through the setup of AWS Clean Rooms and try its different features.

Scenario

The dataset we use is a modified version of the Customer loyalty program data from Northern Lights Air (NLA), a fictitious airline based in Canada. The original dataset is downloaded from Maven Analytics

In this lab, we will set up the loyalty program member and flight activity databases in 2 different AWS accounts.

By using AWS Clean Rooms, the lab will showcase how data analysts can utilize data sources from different entities to perform data analysis without compromising data privacy.

Prerequisite

• 2 AWS accounts
• Admin access on each AWS account (Both console and API access)
• Terraform

Setup your environment

To deploy resources efficiently, the Terraform templates in this lab will be using 2 AWS profiles named:

• aws-clean-rooms-lab-account-1

  This is the account hosting members' database.

• aws-clean-rooms-lab-account-2

  This is the account hosting the flight activity database.

You should follow this AWS guideline to set up your local environment and set up these 2 profiles using the admin credentials of each AWS accounts.

You should setup the profiles with the exact names given above.

Terminologies

Complete list of terminologies in AWS documentation: Link

The following are some terminologies we will use in this lab.

• Collaboration

  This is the most fundamental resource in AWS Clean Rooms.

  To start working on AWS Clean Rooms, one AWS account will create a collaboration and invite other accounts to join. All other resources required for analysis collaboration will be created under the collaboration.

• Membership

  When the AWS account joins a collaboration, a membership will link the account and the collaboration.

  This is important because when a data provider grants data access to its data, it's granted to AWS Clean Rooms on behalf of the membership.

• Configured Table

  Configured table represents an AWS Glue table inside AWS Clean Rooms.

  We can set analysis rules on each configured table to restrict data usage over it.

• Analysis Rule

  Analysis rule is the restriction, which we can configure, over what and how queries can be performed over a configured table.

Sessions

0. Prepare Glue database

   This session is to create the AWS Glue databases, which we can use in the lab.

1. Simple Collaboration

   In this session, we will create our first AWS Clean Rooms collaboration with a simple configured table.

2. Joining data from different data sources

   In this session, we will use AWS Clean Rooms to join multiple tables from different accounts.

3. Differential Privacy

   In this session, we will explore AWS Clean Rooms Differential Privacy and experience how differential privacy can be applied to data collaboration.

4. Cryptographic Computing for Clean Rooms (C3R)

   TBC