Introducing Jovo v4: The React for Voice and Chat

Published by Alex Swetlow & Jan König on Nov 18, 2021.

Jovo v4

After 12 months in the making, it's finally here! Jovo v4 improves every step of the workflow, including a new browser-based Debugger, a new CLI, an output template engine, a component-based system, and much more. Learn more below.

Introduction

When we launched Jovo v3 in early 2020, it was a big step towards our vision of enabling companies to build great multimodal experiences that work across devices and platforms. We called Jovo the Voice Layer and added support for voice on platforms like Bixby, the web, and more. Later in 2020, we launched improved support for Jovo for Web.

While v3 was all about expansion (more platforms, more integrations), it was time for us to improve the overall developer experience. We translated everything we learned from 4 years building Jovo into a completely new product.

And here it is! Jovo v4 is a complete rewrite that brings us closer to our vision of building the Multimodal Experience Platform:

  • A truly extensible architecture that can be embedded into your infrastructure
  • A browser-based multimodal testing experience
  • Efficient and integrated workflows powered by the new Jovo CLI
  • A component-based architecture for more structured apps
  • All based on modern TypeScript technologies

Or, if you want to put it like this: If Jovo v3 was the jQuery, Jovo v4 can be seen as the React for voice and chat apps, bringing the development of conversational apps with Jovo to a next level.

All of this was designed and developed in dozens of sessions with the Jovo community. Norbert Gocht from Die Lautmaler shares some insights:

Jovo v4 is a modern component based conversational UI framework with outstanding flexibility and state of the art tooling. We were able to deliver a deeply integrated custom conversation platform and tailored conversation applications for our clients. Thanks to open source, the incredible dedicated team of Jovo and its amazing community - Jovo v4 is a top solution for professional development of all kind of conversational applications.

I'm especially proud of what we accomplished given that we're only a team of 4 people. Huge thanks go out to Ruben, Max, and Alex for this! And to Kaan who has been a lot of help earlier to year. It was a pleasure. And to our community members who helped us with feedback, feature requests, and pull requests. It's great to be part of such a great community!

Can't wait to see and hear what everyone is going to build with Jovo ´v4`!

Jovo v4 Features

Learn more in our getting started guide or in this video:

https://www.youtube.com/watch?v=YwsLarNuMW0
  • New Debugger: An all-new Debugger with multimodal testing capabilities
  • Components: Reusable, isolated elements for a better project structure
  • Output: A multimodal template engine that works across platforms
  • CLI: Rebuilt from the ground up, including plugins and custom commands
  • More: Lots of more improvements and updates

New Debugger

A screenshot of the new Jovo Debugger

The new Jovo Debugger comes with lots of improvements:

  • Test your app by typing, speaking, or pressing buttons
  • See the whole data flow in the Lifecycle view
  • Customize the Debugger with a jovo.debugger.js file

Learn more about the Jovo Debugger here.

Components

A Jovo Framework component that asks the user if they like pizza

Jovo v4 has a component-based system at its core, similar to popular web frameworks like React and Vue. This makes it possible for our users to build reusable elements that can be used across projects.

Learn more about components here.

Output

Jovo v4 comes with a first of its kind multimodal template engine that turns structured output into native platform responses.

For example, the output below gets translated into the appropriate format for all platforms that you can find in the Jovo Marketplace:

{
  message: 'Do you like pizza?',
  quickReplies: ['yes', 'no']
}

It comes with the following features:

  • It makes it easier to separate content from logic
  • Multimodal: Works for voice, chat, and visual output
  • Works across platforms, even on Alexa APL
  • Supports validation and automatically trims content based on platform requirements

Learn more about output here.

CLI

A screenshot of the new Jovo CLI

The CLI was completely rebuilt from the ground up:

  • A new, more modular architecture
  • An improved build and deployment process
  • Build your own CLI plugins and custom commands

Learn more about the Jovo CLI here.

More

Getting Started with v4

There are several ways how you can get started with v4: