Form autofill
In Pega Customer Service™, natural language processing (NLP) analyzes the customer conversation and automatically fills blank fields in a form. For example, a customer calls U+ Bank support and says "My son was born on September 24 and I want to add him to my health insurance." NLP detects the birth date and adds it to the Add child dependents form.
Voice AI automatically fills empty fields. If a field in the form already has an entry, it is not overwritten.
When Voice AI detects an entity such as customer name, or date, the entity is added to the entity count shown to the right of the current dialog. The CSR can click the count icon to display a list of Voice detected values.
The CSR can either approve Voice AI suggestions or overwrite the fields as appropriate before submitting the form. This feature improves the productivity of a CSR by automating labor-intensive, repetitive work.
To detect appropriate entities from the conversation, you need to:
- Make sure one or more NLP entity extraction models are assigned to the channel
- For each case type, map entities to case properties
Case properties
Each case type has associated case properties. For example, the Account address change case includes the properties for city, state, and zip code. The case properties list comes from relevant records. If you need to add a property to use in a case type, you add the property as a relevant record.
In Dev Studio, select Configure > Application > Inventory > Relevant records.
How Voice AI detects an entity and associates it with a case property
To configure Form Autofill, you map entities to case properties. In the suggested case configuration, you add each entity and the associated case property. For example, for the Account address change case, in the Entities extraction section, you map the address entities (street address, city, state) to the appropriate case property.
The extracted entities are used to automatically fill form fields. When an entity is found in the conversation, if it is mapped to a case property, it will auto-fill an empty form field. If field is not empty, will not overwrite the current value.
When the Voice AI channel is created, Pega Customer Services creates a new Text Analyzer, and adds an extraction model. An extraction model contains the entities that Voice AI can detect in a live conversation. A data scientist can view and make changes to entity models in Prediction Studio.
By default, new Text Analyzers include the pySystemEntities extraction model, which contains 20 entities. As Voice AI processes customer conversations, you can add new entities to a model.
This Topic is available in the following Module:
If you are having problems with your training, please review the Pega Academy Support FAQs.
Want to help us improve this content?