Difference between intent and slots

I am struggling a little bit between the diff of intents and slots

Nancy - Intents are tasks your voice assistant can do for you (e.g. turn on the AC, play a song, tell a story).

Slots are optional pieces of info you tell your voice assistant to look for, in order to tweak what you want to do (e.g. turn on the AC on high, play a soothing song, tell a scary story)

Also, have you checked out the tutorials by VoiceFlow?



Yes I have my situation is a little unique. I want to capture a verbal response – as the user is practicing a business conversation. I am trying to use the Intent as a way to decide did they say something close to what I was expecting them to say and then the Slots to look for something very specific in the captured response.

Example: User is instructed to say “Thank you for calling CVS, my name is Nancy. How may I help you today?”

If the user stated something close to this or a close variation, then Good. Also If response contained Thank you and my name and How may I help. Then the bot will responds as a correct response otherwise, the bot will respond – “You didn’t greet the customer properly try again.” And loop them back to the previous step.

Not sure if my approach is the best way to use VoiceFlow

Nancy - Hello, analyzing recordings of users and giving them feedback, for example, on what words they used and in what order - my generalized understanding of your use case - is not something Alexa/Amazon architecture is designed for. This 5 year old stackoverflow thread touches on this:

“Amazon’s Alexa service is not designed for dictation. This has been the consistent response from the Developer Evangelists… Internally Alexa uses the interaction model to build up a tree of possible sentences that might be recognized. It then takes the input sounds, and matches them against the tree. Whichever branch ends up with the highest confidence is the branch that is selected.”

Whether VF can do what you want via another voice platform, e.g. mobile app, @Nicolas & team should probably weigh in since I’ve only worked with Alexa.


Yes I know that Alexa is not able to do this but I think there is a way for Voiceflow to analyze the captured verbal response and determine if the response was what is expected or not using the If Condition

1 Like

In Alexa skill, verbal response will be always captured by Alexa “FIRST”. Alexa will analyze those like which intent users’ uttterances is matched with, or which keywords should be identified as slots, THEN Alexa will pass ONLY those intent name and slot values to Voiceflow.

it is important to know that speech recognition, language processing and understanding are always done on Alexa, not on Voiceflow, as far as speaking about Alexa projects.

1 Like

Voiceflow has not supported yet, but Alexa conversations might be suitable for your use-case in future, I guess.

1 Like

Thanks for sharing this information.