Design Pattern for daily upload of source data i(in graphql) into dynamo

Hello I am looking for best practices in designing a simple solution to retrieve a large data set (100K orders) daily into a dynamo DB.

At the moment, the source data (orders) is accessible via graphql api , but the solution needs to cover other methods, such as as REST API.

Also, I need something configurable the data sources object model evolve on a regular base.

In the old days of EAI, we would simply use an out of the box configurable adapter (usually with object mapping wizards) for mapping source object modes (at data field level) to target object data model. Just wondering what the AWS best practices for this scenario? thanks a lot :)