Is there a way to mark a question as a test when checking if Fin AI agent is responding as expected after changing guidance? I have had members of my team ask various questions after uploading a “source of truth” to the knowledge base to ensure Fin is responding correctly. Some of these are marked as resolved and others are just sitting in our inbox as unassigned, and I don’t want to mark them as resolved if that is going to create a charge for a test rather than a true customer issue resolution.
Answered
Fin AI billing for testing
Best answer by Gerald Prado
Outside of using the testing feature under Fin I think this may be a good place to try using attributes. You can create attributes that categorize the content of the conversations.
You could then use fin to detect a specific attribute that’s in the same realm as the guidance you’ve updated and then assign the conversation a tag for reporting using a workflow, or send the conversation to a specific inbox instead of resolving the conversation.
+2Outside of using the testing feature under Fin I think this may be a good place to try using attributes. You can create attributes that categorize the content of the conversations.
You could then use fin to detect a specific attribute that’s in the same realm as the guidance you’ve updated and then assign the conversation a tag for reporting using a workflow, or send the conversation to a specific inbox instead of resolving the conversation.
Join the Intercom Community 🎉
Already have an account? Login
Login to the community
No account yet? Create an account
Intercom Customers and Employees
Log in with SSOor
Enter your E-mail address. We'll send you an e-mail with instructions to reset your password.