Sdílet prostřednictvím


Enabling Voice Interactions

You can easily enable voice interactions when using our Web-Chat on desktop browsers or when developing your own mobile experience. For more information about the Web-Chat and how to use it please refer to our WebChat GitHub project.

Web-Chat Voice Interactions

When the voice interactions feature is enabled, the Web-Chat's send button changes from "a screenshot of the webchat with chat option" to "a screenshot of the webchat with voice option". To start a voice interaction, the user needs to select the microphone icon. From there on, the bot will automatically switch between listening to the user and reading the bot's next message.

Important: Note that this experience is optimized for desktop browsers and not mobile browsers. Most mobile browsers don't support audio autoplay and thus degrade the voice interactions experience.

Google Chrome

Google Chrome provides built-in support for speech recognition allowing the browser to locally convert speech to text and vice versa. The voice interactions option should be enabled by default.

Microsoft Edge & Others

To enable voice interactions on other browsers such as Microsoft Edge, you'll need to use the Cognitive Services Speech API. After creating a new account for Speech API, copy one of the provided subscription keys and paste it as a query parameter named 'speechApiKey' to the Web-Chat's URL. For example, if your Web-Chat URL is https://www.contoso.com/webchat/index.html , to enable voice interactions it should be https://www.contoso.com/webchat/index.html?speechApiKey=YourSubscriptionKey.

Mobile Native Experience

If you develop your own mobile experience, you can use Cognitive Services Speech API for speech recognition and synthesis. We recommend using Speech API's native SDKs to easily enable voice interactions in your mobile application.

Next steps

Configure interrupting scenarios