Meet the 3CX AI Customer Service Agent

Handle routine support questions before handing over to a human. The new agent goes beyond call routing, listening to the issue, searching your knowledge sources, and giving a useful answer before deciding whether a person is needed.

Unlike an AI Receptionist, which mainly directs callers to the right destination, the Customer Service Agent can listen to the issue, search your available knowledge sources, and provide a useful answer before deciding whether a human agent is needed. It's designed for level 0 and level 1 support tasks, helping teams handle common questions from documentation, FAQs, policies and product guides. This feature requires 3CX AI Edition and V20 Update 10 or above.

Resolve more calls before handoff

An AI Receptionist uses natural language to understand a caller and route the call. The AI Customer Service Agent adds a support-focused flow designed to resolve routine requests before a transfer is needed. The agent can greet the caller and identify the issue, use web search for more information when enabled, ask one question at a time to clarify the request, give a short answer or one to three practical steps, transfer the call when the issue needs human support, and search configured knowledge sources, including FAQs, policies, product guides and support documentation.

Well-structured knowledge sources help the agent find relevant information quickly. Keep each source focused on one topic, and use clear descriptions so the agent can select the right content.

Escalate when a human is needed

The agent aims to resolve the request first. It transfers the call when it can't resolve the issue, when the caller asks for a person, or when specialist support is required.

Admins can set a default human handoff destination, such as a support queue, and configure escalation topics to route specific requests to a human agent, queue, or another AI Customer Service Agent with more specialised knowledge. When a human agent takes over, caller details are available in the Web Client, and the agent can also review the call transcript when more context is needed.

Use it across your support operation

Customer Service Agents can support several common call flows:

• First-line support: answer routine questions before they reach the support queue
• Queue overflow: help callers when the support team is busy
• After-hours support: provide answers outside normal service hours
• Specialist routing: transfer specific topics to the right person, queue, or specialised AI Agent

For better results, use several focused agents instead of one agent designed to cover every possible scenario.

Configure your customer service agent

Before creating the agent, prepare its knowledge sources and decide where unresolved calls should go. Use the latest recommended AI models for the best available performance.

If you currently use an AI Receptionist for customer service requests, move these tasks to the dedicated AI Customer Service Agent. The Receptionist can focus on routing while the Customer Service Agent follows a support-focused resolution and handoff flow.

Got questions?

If you'd like help setting up an AI Customer Service Agent, choosing the right knowledge sources, or deciding where escalations should go, get in touch with your Zen Systems account manager or reach our support team directly.