The AI-powered Augmented agent improves the conversational experience between customers and customer service representatives (CSRs) by demonstrating the human traits of emotional intelligence in conversations. Though the CSR remains at the core of the customer service experience, augmented agents make it easier for customers to quickly find relevant information. The Augmented Agent simulates human conversation by suggesting relevant replies to the customer service representative (CSR) in response to customer input during a live chat. The agent analyzes and matches the text pattern of customer input with lists of available common phrases, knowledge articles, and page push entries, and then suggests some of the most probable replies to the CSR.
The following figure illustrates a use case for Augmented agent, displaying a suggested reply to a customer request during a live conversation:
Pega Intelligent Virtual Assistant (IVA) is the chatbot technology at the heart of the Augmented Agent feature; IVA interacts with an application by sending and receiving text entered by users. When the customer sends a request in a chat session, the IVA or bot logic captures and routes the customer request to Customer Decision Hub (CDH). CDH predicts customer behavior during live interactions and responds accordingly to the customer.
CDH packages Natural Language Processing (NLP) models under a text analyzer rule. NLP takes customer interaction and processes it for:
- Sentiment analysis
- Text (topic) classification
- Intent analysis
- Entity extraction
Customer Decision Hub uses Adaptive Models to give a score to every relevant suggestion that the Augmented Agent makes to the CSR, and then adapts that prediction. A Topic Model then analyzes the customer’s utterances and automatically classifies the customer queries into appropriate categories. In the absence of Customer Decision Hub, the augmented agent receives suggested replies from the topic models.
The figure above shows how CDH predicts the customer’s behavior and suggests a probable reply to their request, using the Suggested replies option. For example, when the customer asks, Can I pause my service temporarily?, the text analyzer looks for an appropriate topic model and suggests the Pause Service topic from the Common phrases Adaptive Model. Based on the customer utterance, the Augmented Agent suggests the most probable response to the CSR. The CSR can then send the suggested response to the customer.
You can use Customer Decision Hub with Pega Customer Service to drive upsell, cross-sell, and retention offers and provide additional guidance to representatives on the next best action for the current customer and context. However, enabling the Customer Decision Hub settings in Pega Customer Service is no longer mandatory to receive suggested replies and next best actions.