Summarize PDF and .docx by ChatGPT, stored in Google Sheets and sent by email

What are you trying to achieve?

My goal is to create an automated workflow that :

  • takes articles or documents uploaded by a user on Google Drive,
  • picks a title or generates one,
  • generates relevant summaries and SEO-optimized keywords using ChatGPT,
  • saves the results in a Google Sheet.
    As a final step, this information must be summed up in an email once every 3 documents.

If a google doc must be created in the process, I want it to be deleted after the email is sent to that the Google Drive is not cluttered

Thanks for your help :slight_smile:

Steps taken so far

So far, I created a process including a router, but:

  • I have an error message with the pdf module: “The operation failed with an error. Async Job Error: [200] Input document is damaged or of incorrect type. If you use link to Google Drive, Google Docs, Dropbox or similar: they may start returning empty page instead of your file when too many requests are made. If you have this issue, please try PDF.co built-in file storage located at PDF.co instead. If you want to use Google Drive or similar anyway, please try to increase delay between calls to this file”
    I checked my pdf and it is readable and I would prefer to store my documents on Google Drive rather than PDF.co storage.
    Plus, isn’t there any free way to parse .pdf document without having to use API key or any sort of credit?

  • I don’t know how to handle .docx extension document. I guess I have to convert them into Google Docs? This is why I added the “create a document” and “get content of a document” Google doc modules. I didn’t find any Word modules dedicated to the parse task for my use case.

  • How can I set up the rule “send an email with summary every 3 documents” please ? I guess by an array filter but I could use some help here

Thank you very much

Screenshots: scenario setup, module configuration, errors

Blueprint - pdf docx to summarize.json (166 KB)