Y Meadows NLP (Natural Language Processing) models are trained to interpretspecific message or request types, which we call ‘Intents’ (i.e. Password ResetRequest or Merchandise Return Request).

Creating new intents is done by training the AI Model using a set of data (or sample messages) that have intents already categorized. The process of creating the data set is done in the Labeler. Users can quickly select and label each message.

In the example below, a user in Underground Airlines is labeling messages using 4 different intents: 

1. Lost Luggage

2. Missing Points

3. Customer Complaint

4. Other (any message that doesn't belong in the other 3 categories)

‍

‍

The labeled messages are automatically saved and ready to be used to train the AI model. It only takes a few minutes to accurately label and train a Y Meadows model.