VOICE AGENTS
An example of a voice agent that does real work, not just answers questions.
AI voice agents that communicate naturally with users and connect conversations to real actions through tools and automation.
Process snapshot
Illustrative view of how this system is structured — workflow characteristics, not measured client results.
- Workflow views
- 02
- Voice automation
- Tool-enabled
- Actions
- Appointment
Workflow views
Voice automation
Actions
Automation coverage by stage
- 01
A conversation or webhook event reaches the agent
- 02
Fields such as locations, dates and intent are extracted
- 03
The AI agent decides which connected tool to call
- 04
Tools handle lookups, messaging and appointment actions
Bars indicate how far each stage runs without manual input — illustrative, not measured.
Overview
The agent handles the conversation, while an automation layer behind it performs the actual work: reading and writing records, sending messages and managing appointments.
A dedicated MCP server exposes appointment-management tool workflows to the voice receptionist, so the agent can act during a conversation rather than only answering questions.
The challenge
Routine phone and voice requests — availability checks, bookings, reschedules — occupy a person full time even though each request follows the same few steps.
The solution
A voice agent that handles the conversation while an automation layer executes the actual actions through connected tools during the call.
How it works
- 01A conversation or webhook event reaches the agent
- 02Fields such as locations, dates and intent are extracted
- 03The AI agent decides which connected tool to call
- 04Tools handle lookups, messaging and appointment actions
- 05Results are written to Google Sheets and confirmed back to the user
- 06Email and WhatsApp are used for follow-up where relevant
Automation & architecture
Tool-enabled appointment system — MCP appointment tools. Each component is wired so data passes forward in a predictable shape, with the AI layer making the decisions and the automation layer carrying out the actions.
- Natural voice interaction
- AI agent processing and routing
- Connected tool execution
- WhatsApp interaction
- Gmail and Google Sheets integration
What it automates
- Understanding spoken requests and extracting details
- Choosing and calling the right connected tool
- Checking availability and offering alternate slots
- Booking and rescheduling appointments
- Writing results back to Google Sheets
- Following up over email or WhatsApp
My role
Built and configured the voice agent, connected external tools, and tested real-time tool execution.
Benefits
- Handles conversations naturally
- Reduces manual handling of routine requests
- Connects conversations directly to real actions
- Automates appointment-related tasks
- Reaches connected systems through tools
- Provides a more direct conversational experience
Have a process worth automating?
These are examples of systems I've designed — the approach adapts to most business workflows.