Skip to content

Amazon Lex Intents - Capture any utterance

0

Is there any way to configure an Amazon Lex Bot intent where everything that the user says as input is recognized as this intent?

I imagine something like a wildcard utterance? For example could be "*", meaning that everything the user is saying is going to be identified as the intent configured and not failing to the fallback intent.

My requests come from a project where I need to be able to pick customer voice inputs in Amazon Connect and send it through a lambda function. But sometimes what the user says is not picked up by the intent i configured and is failing to fallback intent and there I need to create a loop that again sends the user to ask for the input again.

asked 2 years ago842 views
2 Answers
1
Accepted Answer

Just add a fulfillment lambda block to your fallback intent (if you are using the visual builder), or select the button to set fulfillment to "active" if you are using the standard editor.

AWS
answered 2 years ago
EXPERT
reviewed 2 years ago
  • Thank you Thomas, it's working!!

0

No. However what you can do is configure a fulfillment lambda that takes the transcribed utterance regardless of the intent and stores it in a session attribute. You could then access this lex session attribute in Amazon Connect. You could also create a dummy intent that won't get triggered by the customer's utterance and just handle everything through the fallback.

AWS
answered 2 years ago
  • Thank you Thomas, you are right that is a workaround for what I need, what I have not been able to do is to make that the fallback intent activates the fullfilment lambda. Is there any instructions that you may have or provide me? Because i only get this Lex Session Attributes returned by the fullfilment lambda once the intent i configured as "CaptureAllIntent" is triggered. My solution would be great if I could manage everything by the fallback intent. Could you help me with that please?? I'm super grateful with your insights.

You are not logged in. Log in to post an answer.

A good answer clearly answers the question and provides constructive feedback and encourages professional growth in the question asker.