How to extract structured data from documents for document merge with N8N?

Hi everyone!

I’m a lawyer, and part of my work involves creating similar legal documents (templates) repeatedly.

I’d like to build a workflow where I can process multiple documents, extract key fields (like name, address, amount, date, etc.) — ideally into a JSON file — and then use N8N to run a merge that fills these fields into a Google Docs template (for example, a formal notice letter).

Is it possible to achieve this with N8N, maybe using GPT or another tool for the extraction?

And how can I automate filling a Google Doc with these fields from the extracted data?

Thanks a lot for any advice!

10 comments