To create a new custom entity, go to datasets, open the dataset in the language where you want the entities to be recognized and click custom entities in the menu on the left and then create new entity.
The entity editor will open and you'll be able to start working on your entities. First, you will have to name the entity. Think of it as a category of things. In our example we'll use phones as category and "iphone", "samsung galaxy", etc as values.
Write the name of the entity in the field at the top and click V to confirm
Now you are ready to specify the values that the system should recognize as entities and the synonyms that should be considered to be part of the same value.
Add as many values and synonyms as you deem fit for your case, doing so it will greatly increase the accuracy of entity detection.
Repeat the process for all the custom entities that you need to train and once you are done click SAVE.This will trigger entity training for your test bot.
Once you are sure that the entities are working as expected, click push live and add entity support to your live bot.