1. Home
  2. Docs
  3. Publishers
  4. Basics
  5. Intent and Entity model

Intent and Entity model

Meet the Convoworks intent model which is automatically propagated to target platforms


One of the most widespread ways to handle conversations is an intent-based approach. It is used by Alexa, Dialogflow, Wit.ai, IBM Watson, and many more platforms. The intent-based approach is focused on determining entities and intents of natural language sentences. 

Basically what you have to do is define user utterances (what the user might say) and you group them in intents. During this process, you can also mark particular words you would like to extract as entities.

We’ve got the same thing here in Convoworks, with the addition that we can automatically translate Convoworks model to a particular targeted platform.

We recommend checking the official Amazon and Dialogflow intent models.

System intents and entities

System intents and entities are provided through Convoworks extension packages. They are defined in a cross-platform manner so they can be automatically be propagated to the targeted platforms.

Convoworks system intents are sometimes mapped onto the platform system intent and sometimes are mapped onto the platform custom intent.

Here is an example of convo-core.YesIntent which will be translated to:

  • AMAZON.YesIntent – system intent on Amazon
  • YesIntent – custom intent on Dialogflow defined with utterances: yes, yup, sure, cool …

You can see the full list of available system intents here and the system entities here.

Custom intents and entities

Custom intents and entities are defined at the service level and you define them to fit your service’s needs.

Intents and entities are stored in such a manner that we are able to translate them to target platforms (Alexa and Dialogflow so far).

When you are defining custom intent utterances you can use both custom and system entities to single out significant utterance parts.