Problem: Alexa Skill Builder & Google Dialogflow need separate back-ends
Modern voice assistants, such as Google Home or Alexa, are gaining ground at the moment. Both systems are based on a similar approach: interaction models are created within the individual skills (for Alexa) or actions (for Google Home). An interaction model depicts different ways for possible dialogues between user and voice assistant and which information can be extracted from it.
For this purpose, Amazon provides the “Skill Builder”, which can be used exclusively for Alexa. The counterpart from Google is called “Dialogflow” (formerly API.ai) and has a number of other integration options in addition to Google Home. With the help of “Intents”, the user requests are forwarded to a back-end for further processing. Each platform offers its own SDK with which the respective “Intents” can be processed. This means, a separate back-end has to be written for each platform, which makes both development and maintenance work much more difficult.
Jovo! A developer platform specifically designed for the development of Alexa skills and Google Home actions.
Jovo has several useful features:
- Jovo CLI: a tool to design, build and deploy projects. (Supports Staging!)
- Jovo Webhook: a free service that creates a link to the local web server. This allows applications to be developed and tested locally.
- Jovo Debugger: a debugger for testing applications using the browser.
- Jovo Language Model: a language model which can be converted into the interaction models of the respective platform.