Using A.I models in Power Automated Flows

Do you want to divide the received emails between the good and the bad ones but you are receiving a lot of emails and do not have the time to read through all of them? Then the Sentiment A.I model might be a good method for you.

Last week I was approached for an automated method to divide the good the bad and the neutral emails received in an email address in order to automate some processes based on the content of the email received, that I am willing to share how I managed to achieve that.

We can start by creating an Instant Cloud Flow that is triggered when an email arrives in the designated email address :

You can add the email address needed for this case. Importance of the email to be read, any filter regarding the attachments or not and of course if we need any subject filtering.

After the email received, we convert the html content to text and we can also concatenate in the method that we like.

Here we use the SentimentAnalysis model that reads through the content of the email by putting that content in the Text field above and gives us an indication if the email is positive or negative.

Last remaining part is setting up the flow to write in a Teams Group.

As seen in the picture above, I am posting as my user but other methods can be used as well. You need to add the Channel, Team Post in and some information regarding the Message that needs to be included. I am using the Global sentiment to receive the outcome of the test email that I will be sent.

This is a sample of the email i am sending to the designated email address and as we can see, is a negative email since we are facing ' login issues'.

As we can see the A.I calculated the negativity in the email received which can help us filter these emails and trigger other functions based on other scenarios that we might have, like create case, opportunities etc. But until then, happy flowing!