AI Voice Agents Challenge - Starter Repository

Welcome to the AI Voice Agents Challenge by murf.ai!

About the Challenge

We just launched Murf Falcon – the consistently fastest TTS API, and you're going to be among the first to test it out in ways never thought before!

Build 10 AI Voice Agents over the course of 10 Days along with help from our devs and the community champs, and win rewards!

How It Works

• One task to be provided everyday along with a GitHub repo for reference
• Build a voice agent with specific personas and skills
• Post on GitHub and share with the world on LinkedIn!

Repository Structure

This is a monorepo that contains both the backend and frontend for building voice agent applications. It's designed to be your starting point for each day's challenge task.

falcon-tdova-nov25-livekit/
├── backend/          # LiveKit Agents backend with Murf Falcon TTS
├── frontend/         # React/Next.js frontend for voice interaction
├── start_app.sh      # Convenience script to start all services
└── README.md         # This file

Backend

The backend is based on LiveKit's agent-starter-python with modifications to integrate Murf Falcon TTS for ultra-fast, high-quality voice synthesis.

Features:

• Complete voice AI agent framework using LiveKit Agents
• Murf Falcon TTS integration for fastest text-to-speech
• LiveKit Turn Detector for contextually-aware speaker detection
• Background voice cancellation
• Integrated metrics and logging
• Complete test suite with evaluation framework
• Production-ready Dockerfile

→ Backend Documentation

Frontend

The frontend is based on LiveKit's agent-starter-react, providing a modern, beautiful UI for interacting with your voice agents.

Features:

• Real-time voice interaction with LiveKit Agents
• Camera video streaming support
• Screen sharing capabilities
• Audio visualization and level monitoring
• Light/dark theme switching
• Highly customizable branding and UI

→ Frontend Documentation

Quick Start

Prerequisites

Make sure you have the following installed:

• Python 3.9+ with uv package manager
• Node.js 18+ with pnpm
• LiveKit CLI (optional but recommended)
• LiveKit Server for local development

1. Clone the Repository

git clone <your-repo-url>
cd falcon-tdova-nov25-livekit

2. Backend Setup

cd backend

# Install dependencies
uv sync

# Copy environment file and configure
cp .env.example .env.local

# Edit .env.local with your credentials:
# - LIVEKIT_URL
# - LIVEKIT_API_KEY
# - LIVEKIT_API_SECRET
# - MURF_API_KEY (for Falcon TTS)
# - GOOGLE_API_KEY (for Gemini LLM)
# - DEEPGRAM_API_KEY (for Deepgram STT)

# Download required models
uv run python src/agent.py download-files

For LiveKit Cloud users, you can automatically populate credentials:

lk cloud auth
lk app env -w -d .env.local

3. Frontend Setup

cd frontend

# Install dependencies
pnpm install

# Copy environment file and configure
cp .env.example .env.local

# Edit .env.local with the same LiveKit credentials

4. Run the Application

Install livekit server

brew install livekit

You have two options:

Option A: Use the convenience script (runs everything)

# From the root directory
chmod +x start_app.sh
./start_app.sh

This will start:

• LiveKit Server (in dev mode)
• Backend agent (listening for connections)
• Frontend app (at http://localhost:3000)

Option B: Run services individually

# Terminal 1 - LiveKit Server
livekit-server --dev

# Terminal 2 - Backend Agent
cd backend
uv run python src/agent.py dev

# Terminal 3 - Frontend
cd frontend
pnpm dev

Then open http://localhost:3000 in your browser!

Daily Challenge Tasks

Each day, you'll receive a new task that builds upon your voice agent. The tasks will help you:

• Implement different personas and conversation styles
• Add custom tools and capabilities
• Integrate with external APIs
• Build domain-specific agents (customer service, tutoring, etc.)
• Optimize performance and user experience

Stay tuned for daily task announcements!

Documentation & Resources

• Murf Falcon TTS Documentation
• LiveKit Agents Documentation
• Original Backend Template
• Original Frontend Template

Testing

The backend includes a comprehensive test suite:

cd backend
uv run pytest

Learn more about testing voice agents in the LiveKit testing documentation.

Contributing & Community

This is a challenge repository, but we encourage collaboration and knowledge sharing!

• Share your solutions and learnings on GitHub
• Post about your progress on LinkedIn
• Join the LiveKit Community Slack
• Connect with other challenge participants

License

This project is based on MIT-licensed templates from LiveKit and includes integration with Murf Falcon. See individual LICENSE files in backend and frontend directories for details.

Have Fun!

Remember, the goal is to learn, experiment, and build amazing voice AI agents. Don't hesitate to be creative and push the boundaries of what's possible with Murf Falcon and LiveKit!

Good luck with the challenge!

---

Built for the AI Voice Agents Challenge by murf.ai