Skip to content

Latest commit

 

History

History
77 lines (57 loc) · 2.75 KB

STREAM.md

File metadata and controls

77 lines (57 loc) · 2.75 KB
Raw

Piopiy API - Python SDK: Streaming Audio Feature

Overview

The Piopiy API allows you to leverage powerful real-time audio streaming via WebSocket during voice calls. This feature is critical for use cases like live monitoring, real-time transcription, and AI-powered voice analysis.

Features

  • Real-Time Audio Streaming: Stream live call audio via WebSocket.
  • Flexible Listen Modes: Choose to stream audio from the caller, callee, or both sides.
  • Configurable Audio Quality: Select from various audio quality options (8000, 12000, 16000 Hz).
  • Stream on Answer: Optionally start streaming only after the call is answered.

Installation

First, ensure you have the piopiy package installed:

pip install piopiy

Usage Example

Here’s how to set up audio streaming during a call using the Piopiy Python SDK:

from piopiy import Action, RestClient, StreamAction

def main():
    # Create instances for actions and streaming
    action = Action()
    stream = StreamAction()

    # Define streaming parameters
    action.stream(
        'wss://your-websocket-url/webhook/stream',
        {
            'listen_mode': 'callee',  # Options: 'caller', 'callee', 'both'
            'voice_quality': 12000,  # Audio quality: '8000', '12000', '16000'
            'stream_on_answer': True  # Start streaming only when the call is answered
        }
    )



    # Example of making a call with streaming enabled
    piopiy = RestClient("your_app_id", "your_app_token")
    result = piopiy.voice.call(
        9198333333, 91898989,stream.PCMO(),
        {
            'timeout': 40,
            'loop': 2,
            'duration': 80,
            'ring_type': 'group'
        }
    )
    print(result)

if __name__ == '__main__':
    main()

Streaming Method Parameters

  1. stream(url, options)
    • url (String): The WebSocket URL where the audio should be streamed.
    • options (dict): Configuration settings:
      • listen_mode (String): Determines which side’s audio to stream. Options are 'caller', 'callee', or 'both'.
      • voice_quality (String): The audio quality in Hz. Options are 8000, 16000.
      • stream_on_answer (bool): Whether to start streaming only when the call is answered (default is True).

Example Use Cases

  1. Real-Time Transcription: Stream call audio to a transcription service for live transcription.
  2. Live Monitoring: Use WebSocket to stream live audio for call quality assurance.
  3. AI-Powered Conversational Analysis: Integrate the streaming feature with AI models for sentiment analysis, customer interaction analysis, and real-time decision-making.
  4. Voice Biometrics and Verification: Stream audio for live voice verification or biometric analysis.