Input and Output data schema

[djmbritt]

One big things that is still missing is defining the input and output schema of the data.
This is important because it will allow us to validate that the ata is correct.

It will also allow users to be able to figure out on the fly when creating batches that the input data is correct.
It will also allow us to make sure that input data from tasks is correct.

For now we should not enforce it when creating a campaign, but for the upcoming release we shoud incorporate it in the roadmap.