📞 AriLink

-powered telephony management with speech recognition and transcription

---

📋 Overview

AriLink is a telephony management system built on Asterisk's ARI (Asterisk REST Interface). It provides voice call handling, transcription, and PBX control capabilities. The system combines WebSockets, RTP, and speech-to-text integration to create a modern, feature-rich telephony solution.

✨ Key Features

• 🔄 Call Management - Handles incoming and outgoing calls through Asterisk PBX
• 🎙️ Speech-to-Text - Real-time transcription using local AI models (Parakeet, Whisper) or Google Cloud
• 🌉 Bridge Management - Creates and manages voice bridges for connecting multiple channels
• 👥 Contact Recognition - Supports voice-activated dialing using a contacts database
• 📡 External Media Channels - Supports external media integration for advanced use cases
• 🔌 WebSocket Interface - Provides real-time updates and control via WebSockets
• 🔁 Automatic Fallback - Seamlessly switches to backup transcription services if primary fails

🏗️ Architecture

The system is built on TypeScript and Node.js with a modular architecture supporting multiple concurrent calls:

flowchart TD
    A[📞 Incoming Call 1] --> B[CallSessionManager]
    C[📞 Incoming Call 2] --> B
    B --> D[Session 1: Bridge 1]
    B --> E[Session 2: Bridge 2]
    D --> F[External Media 1]
    E --> G[External Media 2]
    F --> H[🎙️ Transcriber]
    G --> H
    H -->|routed by session ID| D
    H -->|routed by session ID| E

Loading

Core Components

🎮 AriControllerServer

The main controller that interfaces with Asterisk PBX:

• Manages call flows, bridges, and DTMF input
• Handles Stasis application events (start, end)
• Provides WebSocket server for client connections
• Manages contact lookups for voice-activated dialing

🔤 AriTranscriberServer

Provides real-time speech transcription:

• Connects to configurable transcription services (local or cloud)
• Processes RTP audio streams
• Transmits transcription results via WebSockets
• Supports customizable language and model settings
• Automatic fallback to backup services on failure

📡 RTP UDP Server

Handles the real-time audio streaming:

• Processes incoming RTP packets from Asterisk
• Handles audio format conversion for transcription

🗣️ Transcription Providers

Multiple transcription backend support:

• Local providers: Parakeet TDT, Whisper (runs on your GPU)
• Cloud provider: Google Speech-to-Text API (optional)
• Handles streaming transcription with automatic restarts
• Manages audio chunking for optimal performance
• Provides both interim and final transcription results
• Automatic failover between services

⚙️ Configuration

The system uses environment variables for configuration, including:

Category	Variables
PBX	PBX IP address, login credentials
WebSocket	Server ports, external host information
Transcription	Language settings, model configuration
Telephony	Provider settings, phone numbers

🚀 Getting Started

Prerequisites

1. Set up FreePBX server:
   • 📦 New installation? FreePBX Installation Guide - VM setup and FreePBX installation
   • ⚙️ Already installed? FreePBX ARI Configuration - Configure for AriLink
2. Install UV (Python package manager):

   # Windows PowerShell
   irm https://astral.sh/uv/install.ps1 | iex

Installation

1. Setup Transcription Service (local speech recognition):

   Parakeet (Recommended - fastest):

   cd transcription-services/parakeet-service
   uv venv
   uv pip install -r requirements.txt

   OR Whisper (alternative):

   cd transcription-services/whisper-service
   uv venv
   uv pip install -r requirements.txt

2. Configure environment variables in .env file:

   TRANSCRIPTION_SERVICES=ws://localhost:5000

   See .env.example for all options and fallback configuration.

3. Configure contacts in tools/contacts.json for voice-activated dialing

Running the System

1. Start Transcription Service (in terminal 1):

   For Parakeet:

   cd transcription-services/parakeet-service
   start-service.bat

   OR for Whisper:

   cd transcription-services/whisper-service
   start-service.bat

   First run will download the model (~800MB for Whisper, ~600MB for Parakeet)

2. Start AriLink Server (in terminal 2):

   npm start

See Transcription Services Guide for all configuration options including fallbacks.

💡 Use Cases

📞 Voice Call Center

Handle incoming calls with transcription for record-keeping

🤖 Automated Calling Systems

Set up outbound call campaigns with speech recognition

🗣️ Voice-Activated Dialing

Allow callers to speak names instead of dialing numbers

📝 Call Recording with Transcription

Keep searchable records of call content

📦 Dependencies

• Asterisk PBX with ARI enabled
• Node.js and TypeScript
• Transcription Service - choose one:
  • Local: Parakeet TDT 0.6B (RECOMMENDED) or Whisper
  • Cloud: Google Cloud Speech API credentials (optional)
• Various NPM packages including:

  ari-client, @google-cloud/speech, ws, express, dotenv
  

🔮 Future Improvements

• 🔧 Enhanced typing for TypeScript
• 🖥️ Web UI for monitoring and management
• ✅ Additional speech recognition providers DONE: Local Whisper model integrated!
• 🔊 Additional providers: Scribe from ElevenLabs, Azure Speech
• 📊 Call analytics and reporting features
• 💾 Database persistence for call records and transcriptions

📜 License

This project is licensed under a Non-Commercial License (MIT-Based) - see the LICENSE file for details.

Summary

• ✅ Free for non-commercial use - Use, modify, and distribute for personal and educational purposes
• 💼 Commercial use requires permission - Contact for commercial licensing
• 📧 Get in touch: Discord: alexispace

Key Points:

• The software is provided "as is" without warranty
• Attribution is required in all copies
• Commercial use requires explicit permission from the copyright holder

For the full license text, please refer to the LICENSE file.

---

Built with ❤️ for modern telephony solutions