Skip to main content

Struggling to summarize the text of the zoom meeting using ChatGPT zap. The first setup was:

  • New meeting in Zoom zap
  • ChatGPT Summarization, pass in as the action the trans_dl_urls or the trans_play_urls

When I did this the text it sends to chatGPT isn’t the text, it a link:

system_message

Please summarize the user's provided text. Use a method of Extreme TLDR generation, a new form of extreme summarization for paragraphs. TLDR generation involves high source compression, removes stop words and summarizes the paragraph whilst retaining meaning. The result is the shortest possible summary that retains all of the original meaning and context of the paragraph.

text_to_summarize

https://....zoom.us/...

summary

The user provided a link, but no specific text.

I tried using Google Docs:

  • New Zoom meeting
  • Save Google Doc in Folder
  • New file in google drive folder
  • Pass the ‘File Txt’ from the new google drive folder zap into ChatGPT

That gives this error though:

system_message

Please summarize the user's provided text. Use a method of Extreme TLDR generation, a new form of extreme summarization for paragraphs. TLDR generation involves high source compression, removes stop words and summarizes the paragraph whilst retaining meaning. The result is the shortest possible summary that retains all of the original meaning and context of the paragraph.

text_to_summarize

https://zapier.com/engine/hydrate/3541577/…./

summary

The provided text is a URL, not a paragraph. Please provide a paragraph for summarization.

Which I read up on hydrate and I guess it’s doing this to make stuff faster, but I can’t figure out how to ‘hydrate’ it when it passes it into ChatGPT. Thanks for any help you kind folks can provide.

Hi @Extole 

Good question.

To help us have true context, post detailed screenshots with how your Zap steps are configured.

Try using this Zap trigger: Zoom - New Recording

 


Here @Troy Tessalone hope this helps:

 

These are for the Google Docs attempt:

 

This was directly from Zoom->ChatGPT:

 


@Extole 

 Check this related topic for how to get the contents of a file.

 

 

Both of your examples are providing files (file object or file url) to a field that expects text.


 


Thanks @Troy Tessalone , I don’t seem to get a ‘text’ output from the GET Webhook, here is what I configured and the test results:

 


@Extole 

We would need to see screenshots with how your Zap step 2 is configured.


Sorry, I thought I had posted that, but here they are. Thanks so much for your help, would love to buy you several coffees if there is a way to do that (:

 


@Extole 

What type of file are you sending to Step 2? (e.g. .txt, .docx, etc.)

Can you show screenshots of the DATA OUT from Step 1 related to the file type?

 

Step 2 should return the text if configured correctly, which you would then pass to Step 3.

 


Ah thank you, here is maybe the problem. If I look at the Zoom Zap it has these fields:

I was trying to store the ‘trans_dl_urls’ in google docs, (without the webhook thats there now). That it turns out was just giving me a google doc that had just that URL in it, not the chat transcript. Per your suggestion above I thought maybe I could take that URL and instead use the GET Webhook:

testing here:

 

Which now gives me a google doc with instead of the zoom download link, a hydrate link like this:

'https://zapier.com/engine/hydrate/3541577/.eJwtzc1qAjEUhuF7OeuJnfyMcdy5arsoSCm4KCXE5ETCaDIkGYsMc-9Gcft-h-fMYGIoGIoqtxFhC5_QgA-56GBQeV…

Which if I load does show the chat transcript but I still can’t seem to get the text from that transcript URL ): I know I’m doing something stupid here but don’t know why.  ):

One weird thing here might by that Zoom Transcript Download URLs make the browser autosave a .vtt file, which maybe isn’t text so it gets confused? Thanks again for all your help, really appreciate it!

 

 


@Extole 

You need to use the true file download URL.

So if that link is within another file, then you will need to parse it out.