Handling AI-Powered Conversations
The AI Agent enhances live chat by handling customer queries with automated responses. Once enabled by an admin, it operates 24/7, even outside business hours, ensuring continuous support.
How the AI Agent Interacts with Customers
When the AI Agent responds to a customer query, the following options may appear:
- Talk to a Person – Reassigns the conversation to a live agent.
- That Helped – Marks the conversation as resolved and closes it.
Note: These options are not shown for basic interactions such as greetings, thank-you responses, or welcome messages.
Handling Reassignments
Customer Escalation
- If a customer selects Talk to a Person, the conversation is immediately reassigned to a human agent (if available).
- If a customer selects That Helped, the conversation is marked as resolved and closed.
Offline Hour Handling
If no live agents are available, the AI Agent manages escalations as follows:
- Offline Message – An automated message informs the customer that agents are unavailable.
- Offline Form (if enabled) – The customer receives a “Contact Us” option to submit their details.
- Ticket Creation – Customers can create a support ticket via the “Contact Us” option.
Additional Agent Controls
Overriding AI Responses
Agents can manually take over a conversation by changing the assignee to a human agent.
Note: Agents cannot assign a conversation back to the AI Agent. They can only take over cases where the AI Agent will not respond further.
Handling Customer Confusion
If a customer is stuck with the AI Agent, agents can take over by assigning the conversation to a human agent to provide direct assistance.
Performance Insights
To improve AI responses and understand gaps, agents can review past AI interactions. Refer to the article on Unanswered Questions by AI for more details.
Summary
The AI Agent ensures seamless support through automation, reducing workload on human agents while providing customers with options for escalation when needed. Agents can take over conversations when necessary and leverage performance insights to refine AI interactions.
