Che Voz - AI Voice Assistant Platform

AgentsText to SpeechSpeech to Text
PreviousNext

Real-time voice assistant with multiple Spanish dialects using ElevenLabs conversational AI

AR
Alan RottaSoftware engineer playing with gen AI, passionate about neuroscience
Che Voz - AI Voice Assistant Platform
DemoCodeVideo

Che Voz - AI Voice Assistant Platform

Overview

Che Voz is an innovative AI voice assistant platform that brings natural Spanish conversations to life with region-specific dialects. Built with ElevenLabs' conversational AI, it offers authentic voice interactions for users across Latin America and Spain, automatically suggesting the most appropriate agent based on the user's location.

Key Features

  • Multi-Dialect Support: Five specialized agents covering Argentine (Buenos Aires, Córdoba, Mendoza), Mexican, and Colombian Spanish accents
  • Location-Based Intelligence: Automatically detects user location and suggests the most culturally appropriate voice agent
  • Real-Time Voice Streaming: Low-latency bidirectional audio using WebSocket technology
  • Modern Web Interface: Intuitive dialer interface with visual feedback and agent suggestions
  • Clean Architecture: Well-structured FastAPI backend with domain-driven design principles

How It Works

Che Voz leverages ElevenLabs' Conversational AI API to create natural voice interactions. The system architecture includes:

  1. Frontend: Modern web dialer that captures audio and handles WebSocket connections
  2. Backend: FastAPI server that manages conversations and integrates with ElevenLabs
  3. Location Service: Intelligent agent suggestion based on IP geolocation and browser location API
  4. Voice Processing: Real-time audio streaming with automatic format conversion (PCM16)
┌─────────────┐     WebSocket      ┌──────────────┐
│   Browser   │◄──────────────────►│   FastAPI    │
│   Dialer    │     Audio/JSON     │   Backend    │
└─────────────┘                    └──────┬───────┘
                                          │
                                          │ HTTPS
                                          ▼
                                   ┌──────────────┐
                                   │  ElevenLabs  │
                                   │ Conversational│
                                   │      AI       │
                                   └──────────────┘

Technologies Used

  • ElevenLabs API: Conversational AI for natural voice interactions
  • FastAPI: High-performance Python web framework
  • WebSocket: Real-time bidirectional communication
  • Docker: Containerized deployment
  • Location Detection: IP-based and browser geolocation APIs

Getting Started

# Clone the repository
git clone https://github.com/alnroot/che_voz.git
cd che_voz
 
# Set up environment
echo "ELEVENLABS_API_KEY=your_key_here" > .env
 
# Run with Docker
docker-compose up
 
# Or run locally
pip install -r requirements.txt
python main.py

Visit http://localhost:8000 and try calling:

  • 111: Argentine Porteño agent
  • 222: Mexican agent
  • 333: Colombian agent
  • 444: Argentine Cordobés agent
  • 555: Argentine Mendocino agent

Demo

Experience Che Voz live at https://che-voz.vercel.app/

The platform automatically detects your location and suggests the most appropriate Spanish-speaking agent. Users can override the suggestion by dialing specific agent codes or letting the system choose based on their region.

Future Plans

  • Expanded Dialect Coverage: Adding agents for Chile, Peru, Venezuela, and Spain
  • Custom Voice Training: Allow users to create personalized voice agents
  • Multi-Language Support: Extending beyond Spanish to Portuguese and indigenous languages
  • Enterprise Features: Call analytics, conversation history, and API access
  • Mobile SDK: Native iOS and Android integration

Team

Che Voz was built by a talented team during the ElevenLabs Hackathon:

Acknowledgments

Built collaboratively during the ElevenLabs Hackathon, Che Voz showcases the potential of AI to preserve and celebrate linguistic diversity. Special thanks to the ElevenLabs team for their powerful conversational AI platform and to the Latin American community for inspiring authentic regional voice experiences.