Currently Alchemy can handle datasets which are in perfect Rows and columns.

Like this:

https://s3-us-west-2.amazonaws.com/secure.notion-static.com/e063b578-4d0b-4aaa-845c-8b2ade3851a3/Untitled

It can not understand data where the first row is not the heading row.

Like this:

https://s3-us-west-2.amazonaws.com/secure.notion-static.com/f6a50e77-cd05-4212-8028-9229222028d4/Untitled

We are building a simple workflow to clean out unstructured headers and and arrive at more structured data.