This library lets you pretend you have a horizontally scalable 'wide column' database, and yet use a relational database like PostGreSQL because you don't currently have the volume of data to justify using AWS DynamoDB.

By modeling your data 'as if' it was using a wide-column store, if you ever need to migrate to DynamoDB, it will be super easy. This model is performant on relational data too, so you are not sacrificing anything by starting this way. Just giving you an easy out if you need to scale.

You know, cause you totally will have to one day, right?

That would be great. Ideally this has a driver similar to DynamoDB's AWS driver, so you just need to change packages to use the real thing. We should totally do that.

The end goal is to have a command that allows for table creation in DynamoDB, and a chunk-exporter for data to go into DynamoDB.

Uh....