ringcentral-web-phone is designed for client side and only works with browsers.

This project was originally designed for server and desktop. It doesn't require a browser to run. It could run in browser too.

• Answer inbound call
• Make outbound call
• Speak and listen, two way communication
• Call control features
  • Redirect inbound call to voicemail
  • Ignore inbound call

• browser demo
• node.js
  • answer inbound call
  • make outbound call
  • redirect inbound call to voicemail
  • ignore inbound call
• call supervise
  • supervise an existing phone call and get real time audio stream

yarn add ringcentral-softphone @rc-ex/core

For node.js you also need to:

yarn add ws wrtc

• for node.js, check here
• for browser, check here

import { nonstandard } from 'wrtc'

softphone.once('track', e => {
  const audioSink = new nonstandard.RTCAudioSink(e.track)
  audioSink.ondata = data => {
    // here you have the `data`
  }
  softphone.once('BYE', () => {
    audioSink.stop()
  })
})

The data you got via audioSink.ondata is of the following structure:

{
  samples: Int16Array [ ... ],
  bitsPerSample: 16,
  sampleRate: 48000,
  channelCount: 1,
  numberOfFrames: 480,
  type: 'data'
}

Please note that, you may get different numbers, for example, sampleRate you get might be 16000 instead of 48000.

yarn install
cp .env.sample .env

Edit .env file to specify credentials.

• CALLEE_FOR_TESTING is a phone number to receive testing phone calls. You don't need to specify it if you do not make outbound calls.
• If you have WEB_SOCKET_DEBUGGING=true, then all WebSocket traffic will be printed to console.

• for node.js yarn server
• for browser yarn browser

Make a phone call to the phone number you configured in .env file. The demo app will answer the call and you can speak and listen.

Let's say there is a phone call ongoing between a customer and the call agent. You can use this library to supervise this phone call to get live audio stream. You can analyze the audio stream using some AI algorithm and provide tips to the call agent in real time.

Use this library to supervise an existing phone call to get live audio stream. Translate the audio stream into text by invoking some speech-to-text service. Show the text to the caller and/or callee so they can see live transcription.

You can create a program to make a phone call or answer a phone call and play recorded audio. This is good for announcement purpose. This is also good for quick voicemail drop. Or you can use text-to-speech service to read text to the callee.

• How to create a publish message
• How to forward a call