Hello there, I have a Document object

Why do you want to convert it back to Document?
You can directly update the node as well. simply update it with doc ID

I am preprocessing the markdown results from LlamaParse. What I do is remove repeated information that LlamaParse returns as a result of parsing multipage document (e.g. removing company names which appears in every page of the document - I simply do not want to index the same information).

I also need to implement logic that checks if summary amounts exist or not after preprocessing the LlamaParse result. To do this is I make use of MarkdownNodeParser as a quick and dirty way to split the document into nodes and do very specific checks and amendments on the amounts.

Then after all checks are complete, I bring every nodes back together as 1 Document object and put it through MarkdownElementNodeParser for my rag pipeline.