Configure continuous integration for new repository structure

[gsarma]

CI isn't needed off the bat with the new repo structure because it is mostly raw data. But we should think through what tests we will want as we fill out the gaps and add in more code. We should configure CI again before too long to make sure that things don't get overlooked.

I suspect that most of the code will be for parameter extraction since we will likely house all of the channel metadata, plots, digitized figures, and exported NML files in various public repositories.

[gsarma]

@travs maybe you can look at this?