Hello, The un-nest processor in a prepare visual recipe updates the schema only at design time. If your data changes and new keys appear, you will need to go again in the recipe design view. So the behaviour you describe is expected, as we enforce controls over the dataset schemas when you build from the flow. If you want to have a change in schema every time you run a recipe, you will need to write it in Python or R or Scala. There are examples of how to do it online, for instance https://stackoverflow.com/questions/38231591/splitting-dictionary-list-inside-a-pandas-column-into-separate-columns. Cheers, Alexandre