How should the company meet these requirements with the LEAST amount of application changes?

Explanations:

Amazon Redshift is a data warehouse service, not suitable for transactional data management required for customer information and purchases. Using DynamoDB for these aspects does not meet the compliance requirement of keeping data in each region.

While DynamoDB global tables could facilitate a single product catalog, they would not maintain the regional compliance for customer information and purchases effectively. This approach complicates the architecture unnecessarily as it doesn’t provide the necessary separation for sensitive data.

Using Amazon Aurora with read replicas for the product catalog allows a single, centralized product catalog that can be accessed across multiple regions. Additional local Aurora instances in each region for customer information and purchases ensure compliance with data residency requirements, minimizing changes to the application.

While Aurora can be used for the product catalog, using DynamoDB global tables for customer information and purchases does not satisfy the requirement to keep this data within regional boundaries for compliance. This adds complexity without fulfilling the necessary regional data separation.