Entities & Variables

The term Entity refers to a piece of organized data that must be extracted from a user's input. It is a parameter that guides the conversation. Entities are often nouns, such as a product or service, a named object or the time and date, etc.

Entities offer context to an intent itself. They allow your bot to fulfill a user's request by describing the intent in more detail.

Take the taxi booking service as an example:
“I’d like to book a taxi for tomorrow
“Cancel my reservation for 5 p.m.”
“My booking number is 5GB0586
“I need to get to London quickly”
“How to use a business voucher

All the data in bold can be treated as entities as long as they are valuable and important to the dialogue's direction.


Highlighted entities in a phrase

Extracting entity values

Automate will automatically extract entities from a phrase, once trained.


Automatic entity matching and extraction

Above, we see that "taxi_type" is understood, being trained on a variety of vehicle types.
The numeral "7" is also extracted to understand the number of passengers.
The time of "today" is also understood by Automate as a system time entity, a in-built entity of Automate.


Variables are created automatically from entities that have been used in the training data.


Example variable types

The names of the variables correspond 1:1 to entities. Variables serve to retain important data across the dialogue.

In the flow, you can indicate what variable you’re expecting to get at a specific moment of the conversation:


Flow asking for a vehicle type in the conversation

And condition the following steps upon different values of your variable:


Directing a user in the flow based on vehicle type selected

For safety reasons, you cannot delete an entity that has a corresponding variable; you can do it once the variable is deleted too.