What it does Jovo is (was) a cross-platform framework for building conversational apps—voice skills for Alexa and Google Assistant, chatbots for Messenger, Instagram, and the web. It used TypeScript decorators and component-based architecture to abstract platform differences, so one LoveHatePizzaComponent could run more or less everywhere.

The interesting bit The “React for Voice” pitch was ambitious: reusable components, an output template engine that translated structured data into voice, text, or visual responses, and a plugin ecosystem. For a while, it genuinely lowered the friction of shipping to multiple assistant platforms simultaneously.

Key highlights

• Component-based architecture with @Intents decorators and BaseComponent classes
• Output template engine for multimodal responses (voice + text + visual)
• CLI scaffolding, local dev server, and browser-based debugger (now self-hosted or replaced with ngrok)
• Integrations with NLU and CMS services; staging and unit testing built in
• All packages remain on npm in their final versions; repos are read-only as of the shutdown

Caveats

• The hosted Debugger service shut down November 2024; local development now requires ngrok or the newly open-sourced debugger
• The Jovo website is offline—documentation survives only in repository files
• No further updates, security patches, or platform API changes will be handled by the original team

Verdict Worth a look if you’re maintaining a legacy voice skill or chatbot and need to understand what you’re inheriting. Not a starting point for new projects unless you’re prepared to fork and self-maintain.