How to work with file attachments

(Samuel) #1

Hi all! Another question for the group here (thank you all so much for your help thus far!).

I have a new project where the user will upload a CSV spreadsheet for the assistant to analyze. Specifically, I want the assistant to analyze the text in three of the columns and return themes and trends. The task itself is just analyzing text data, which GPT 4o is quite adept at doing. This screenshot is what I am seeing when I test the assistant. I don’t understand why it doesn’t recognize the file.

Any tips and resources are greatly appreciated! Thank you!

image
image.png554×813 38.4 KB

(Nathan Stults) #2

Samual, for the assistant to be able to read the text of uploaded files (or any other file that you can access with a URL) you need to add a “Query File” tool to your assistant.

image
image.png864×896 60.2 KB

Once added, the assistant will be able to extract the text and process it via normal prompting.

If you need the assistant to understand images, we also have a Query Images tool that is similar but uses the Vision capabilities of GPT 4 Omni

1 Like
(Samuel) #3

Excellent! This works perfectly! Thank you very much for the tip about the query image tool as well.