Google Drive Image to OpenAI Assistant

How To
, ,
1

blueprint (3).json (23.8 KB)

I can’t seem to make this scenario work.
Reagrdless of what I do the openai assistant module either tells me the url is invalid or that it should end in png, jpg, webp, etc.

image
image2427×712 126 KB

The purpose of this automation is for the openai assistant to check the 5 images I give it, then check the docs I have in its vector store, and then give me titles for these 5 images depending on the content of the image and of the available docs.

Can’t seem to find an answer on the community.

Thank you in advance y’all!

1 Like
2

Hi everyone

Found an alternative =)

The automation works in the following manner:

  • It watches images in a specific folder
  • it uploads it to Cloudinary so that I can get a link that ends in .png
  • it gives it to the OpenAi Assistant (which is trained on multiple docs and on a specific naming convention, this is why I did not want to use the Analyze image module directly.)
  • Finally it changes the name of the specific file.

blueprint (4).json (20.8 KB)

1 Like
3

Hey @Denis1

If you found this worked, you can mark your answer as a solution to keep community healthy and organised.

Regards,
Msquare Automation - Platinum Partner of Make
@Msquare_Automation

2 Likes