Quick Start Guide

Getting Started

Real-Time Voice AI Accelerator

Get your voice agent running with Azure Communication Services, Speech Services, and AI in just a few steps.

Prerequisites

System RequirementsAzure Requirements

• Python: 3.11 or higher
• Operating System: Windows 10+, macOS 10.15+, or Linux
• Memory: Minimum 4GB RAM (8GB recommended)
• Network: Internet connectivity for Azure services

• Azure Subscription: Create one for free if you don't have one
• Azure CLI: Install Azure CLI for resource management

Microsoft Learn Resources

• Azure Free Account Setup - Step-by-step account creation
• Azure CLI Fundamentals - Essential CLI commands

Learning Paths

🚀 Quick Start (15 minutes)🏗️ Infrastructure First🔧 Deep Dive

Get up and running fast:

1. Local Development Guide - Complete setup with raw commands
2. Architecture Overview - Understand the system design
3. API Reference - Explore available endpoints

Best for: Developers who want to see the accelerator in action immediately

Set up Azure resources properly:

1. Production Deployment - Infrastructure provisioning
2. Configuration Details - Advanced configuration options
3. Local Development Guide - Connect to your infrastructure

Best for: Architects and teams planning production deployments

Understand the complete system:

1. Architecture Overview - System design and patterns
2. Data Flow Patterns - Processing pipeline architecture
3. LLM Orchestration - AI routing and conversation management
4. Operations Guide - Monitoring and troubleshooting

Best for: Technical leads and teams building custom voice applications

Azure Setup Requirements

Required Azure Resources

The accelerator requires these Azure services for full functionality:

Service	Purpose	Required For
Speech Services	Text-to-Speech, Speech-to-Text	All voice features
Communication Services	Phone calls, WebSocket media	Phone integration
AI Foundry / OpenAI	Conversation intelligence	AI agent responses
Redis Cache	Session state management	Multi-turn conversations
Cosmos DB	Conversation persistence	Analytics, compliance

Quick Azure Setup:

# Clone the repository
git clone https://github.com/Azure-Samples/art-voice-agent-accelerator.git
cd art-voice-agent-accelerator

# Deploy infrastructure (choose one)
azd provision  # Azure Developer CLI (recommended)
# or use Terraform/Bicep directly

Development Approaches

🏃‍♂️ Fast Track🏭 Production Ready🔬 Custom Development

Start developing immediately:

• Goal: Voice agent running locally in 15 minutes
• Path: Local Development Guide
• Infrastructure: Minimal (Speech Services only)
• Best for: Proof of concepts, learning, simple demos

Enterprise deployment preparation:

• Goal: Scalable, secure, monitored deployment
• Path: Production Deployment → Local Development
• Infrastructure: Complete (all Azure services)
• Best for: Production applications, enterprise environments

Extend and customize the accelerator:

• Goal: Build custom voice applications
• Path: Architecture Deep Dive → Local Development
• Infrastructure: As needed for your use case
• Best for: Custom voice solutions, specialized industries

Getting Help

Community & Support Resources

Documentation: - Troubleshooting Guide - Common issues and solutions - API Reference - Complete endpoint documentation - Examples & Samples - Practical implementation examples

Community: - GitHub Issues - Report bugs and request features - GitHub Discussions - Community Q&A - Microsoft Q&A - Official Microsoft support

---

What's Next?

Choose your path above and start building your voice-powered applications! Most developers find success starting with the Local Development Guide to see the accelerator in action immediately.

New to Voice AI?

Check out the Architecture Overview first to understand how real-time voice processing works with Azure Communication Services and Speech Services.