Introduction
Most contact centers answer thousands of calls every day. Before a customer reaches a customer service representative (CSR), there are opportunities for an AI agent to do real, meaningful work: verifying the customer's identity, understanding their issue, or retrieving their account details.
That automated first conversation is handled by a voice agent. In Dynamics 365 Contact Center, voice agents are built in Copilot Studio and connected to the voice channel through a voice workstream. When a customer calls a contact center phone number, the voice agent answers. They engage in a spoken conversation using natural language or keypad input. The agent either resolves the issue or hands the conversation off to a CSR with full context in hand.
What voice agents do
A voice agent operates at the start of every inbound call. Depending on how you configure it, it can:
- Greet and route: Identify what the customer needs and direct the call to the right queue.
- Collect and verify: Gather account numbers, confirm identities, and validate information before a CSR picks up.
- Self-serve: Resolve common requests like account balances, appointment bookings, or order status, without any CSR involvement.
- Transfer with context: Hand off to a CSR or external system, passing the conversation history and collected data through the call.
The sophistication of that experience depends on the choices you make as an administrator and solution designer: which orchestration model the agent uses, how it handles different input types, what languages it supports, and how recording and compliance are configured.
In this module, you start by exploring the orchestration modes that determine how a voice agent processes and responds to customer input. From there, you look at the customization options that shape the agent's behavior and conversation design. You also cover the security and compliance settings that govern call recording, data handling, and regulatory requirements. By the end of this module, you have a complete picture of what it takes to deploy a voice agent that's effective and enterprise-ready.