mixedvoices 0.1.3

Mixpanel for conversational AI

MixedVoices 🎙️

MixedVoices is an analytics platform for voice agents - think Mixpanel for conversational AI. It helps you track, visualize, and optimize your voice agent's performance by analyzing conversation flows, identifying bottlenecks, and measuring success rates across different versions.

Features

Core Capabilities

• 📊 Flow Visualization: Interactive flowcharts showing conversation paths and patterns
• 🔄 Version Control: Track and compare agent behavior across different iterations
• 📈 Success Rate Analytics: Measure and optimize performance metrics
• 🔍 Conversation Analysis: Deep dive into individual interactions
• 📝 Metadata Tracking: Store and analyze version-specific configurations
• 🖥️ Interactive Dashboard: User-friendly interface for all operations

Dashboard Features

• Project Management
  • Create and manage multiple projects
  • Version control with metadata tracking
  • Easy navigation between versions
• Flow Visualization
  • Interactive flowcharts of conversation paths
  • Success rate indicators for each step
  • Click-through analysis of specific paths
• Recording Analysis
  • Detailed view of individual recordings
  • Transcripts and audio playback
  • Success/failure tracking
  • Path visualization for each conversation
• Upload Interface
  • Easy upload of new recordings
  • Automatic processing and analysis

Installation

pip install mixedvoices

Prerequisites

1. Python 3.8 or higher
2. OpenAI API key (set in your environment variables)

export OPENAI_API_KEY='your-api-key'

Quick Start

Using Python API

import mixedvoices as mv

# Create a new project
project = mv.create_project("receptionist")

# or load existing project
project = mv.load_project("receptionist")

# Create a version with metadata
version = project.create_version("v1", metadata={
    "prompt": "You are a friendly receptionist.",
    "silence_threshold": 0.1
})

# or load an existing version
version = project.load_version("v1")

# Add recording to analyze, by default this is non blocking and runs on a separate thread
version.add_recording("path/to/recording.wav", is_successful=True)

# run in blocking mode
version.add_recording("path/to/recording2.wav", blocking=True, is_successful=False)

Using Dashboard

Launch the interactive dashboard:

mixedvoices

This will start:

• API server at http://localhost:7760
• Dashboard at http://localhost:7761

Technical Requirements

Core Dependencies

• Python ≥ 3.8
• FastAPI
• Streamlit
• OpenAI API access
• Plotly
• NetworkX

Troubleshooting

Common Issues

1. API Connection Errors

   • Verify your OpenAI API key is correctly set
   • Check your internet connection
   • Ensure you're not exceeding API rate limits

2. Dashboard Not Loading

   • Confirm both API server and dashboard ports are available
   • Check if required dependencies are installed
   • Verify Python version compatibility

3. Recording Upload Issues

   • Ensure audio files are in supported formats (.wav, .mp3)
   • Check file size limits
   • Verify storage permissions

Contributing

We welcome contributions! Please follow these steps:

1. Fork the repository
2. Create a feature branch (git checkout -b feature/AmazingFeature)
3. Commit your changes (git commit -m 'Add some AmazingFeature')
4. Push to the branch (git push origin feature/AmazingFeature)
5. Open a Pull Request

Development Setup

git clone https://github.com/yourusername/mixedvoices.git
pip install -e ".[dev]"

License

This project is licensed under the Apache 2.0 License - see the LICENSE file for details.

Roadmap

• Unit Tests
• Async recording analysis
• Support other APIs and Open Source LLMs
• Team collaboration features
• Custom analytics plugins
• Enhanced visualization options

---

Made with ❤️ by the MixedVoices Team