We are building a DLT pipeline and the autoloader is handling schema evolution fine. However, further down the pipeline we are trying to load that streamed data with the apply_changes() function into a new table and, from the looks of it, doesn't see...
