When processing a line of text with a model. The framework should be smart enough to extract all the known entities and use them to skip steps.
For example lets say we have the following steps defined in our JSON
- What would you like to eat?"
- What size would you like that?"
- What would you like on your pizza"
- Where do you like that ordered to?
- How would you like to pay?
If I type "I want a medium mushroom pizza delivered to my house" the bot should be smart enough to skip all these steps and just ask me for my order confirmation based on the entities ["pizza","mushroom","medium","house."] and my user information ["billing", "address"]
In my opinion this is what can make bots more efficient than apps and forms the ability to encapsulate and extract 5 or more UI steps in a single sentence.
When processing a line of text with a model. The framework should be smart enough to extract all the known entities and use them to skip steps.
For example lets say we have the following steps defined in our JSON
If I type "I want a medium mushroom pizza delivered to my house" the bot should be smart enough to skip all these steps and just ask me for my order confirmation based on the entities ["pizza","mushroom","medium","house."] and my user information ["billing", "address"]
In my opinion this is what can make bots more efficient than apps and forms the ability to encapsulate and extract 5 or more UI steps in a single sentence.