Power Platform Community Forum Thread Details

I am currently working on a project that will read your resume and compare it to some guidelines using an agent in copilot studios. To start this project I am just working on reading in the file and summarizing it. In the testing mode for copilot studio I can successfully do this by using a file array JSON String defined by "JSON(System.Activity.ChannelData)" in a topic and then in a flow convert the first attachment from ['OriginalAttachments'] to Binary to make it into a readable form to get text to inject into a prompt. This is done by using the script "dataUriToBinary(first(outputs('Parse_JSON')?['body']?['OriginalAttachments'])?['contentUrl'])" and then calling the "Recognize text in image or document" action on the output.

This file array looks like this:

File Array: {"OriginalAttachments":[{"contentType":"text/plain","contentUrl":"data:text/plain;base64,VGhpcyBpcyBhIHRlc3QgcmVzdW1lLiBJIGhhdmUgZXhwZXJpZW5jZSBjb2RpbmcgaW4gUHl0aG9uIGFuZCBKYXZhLg==","name":"TestResume.txt"}],"attachmentSizes":[{"Value":67}],"clientActivityID":"80d0c9n5nul","enableDiagnostics":true,"testMode":"Text"}

To keep the base64 string shorter I used a small .txt file but I can also extract text from a pdf if it is in the file array.

My problem I am having with the teams channel is that teams does not use the same file array system so I cant access the file the same way. In teams this is the result of file array looks like this:
File Array: {"tenant":{"id":"741bf7de-e2e5-46df-8d67-82607df9deaa"}}

What I need help with is figuring out how to access the files in the teams attachments so I can use the method I used in testing mode to extract the text from the document. Sort of a "File Array" but for teams. I think I need to find a different script similar to JSON(System.Activity.ChannelData) but for the way that teams stores their files.

If anyone has any insight on this it would be greatly appreciated. Thank you

Categories:

General topics