About: Teresa is the modular AI assistant from the future of the past, powered by Google's Dialogflow and a NodeJS backend. This repository contains the backend code.
Features:
- Teresa comes preloaded with these base set of functionalities: Weather reports, web search(Duckduckgo) and small-talk
- Extremely modular: Teresa is designed to be extended - you can write your own webhooks that add additional functionality without cluttering the codebase.
- Audio feedback: Just like most AI assistants Teresa will respond both with text and synthesized speech.
- All you need is a browser: The frontend adapts to the screen of the device, virtually allowing Teresa to be accessed from any device with a browser.
- No shady tactics: Teresa's backend is open-source, so you know exactly what is happening every time you interact with her.
Instances: You can self-host Teresa. Currently, there's an official instance hosted on glitch:
Read Before Self-Hosting:
- A template
envfile has been provided which will store your backend secrets(API keys, App Credentials,etc) - Rename it to
.envand fill in all the secrets before deploying - Change the axios requests in
/public/js/main.jsto call your self-hosted backend
For Contributors:
- The
publicdirectory consists of the front-end - The
middlewaresdirectory is where you should drop your custom backend middlewares - The
templatesdirectory consists of cards/layouts for display on the front-end, which Teresa will optionally return while fulfilling a webhook - The
routesdirectory consists of all the expressJS routers - The
webhooksdirectory contains js files that will each fulfill a particular webhook - If you've added additional functionality to Teresa and would like to submit your additions, I'd greatly appreciate it.
- Simply open a PR describing your changes
Made with ❤️ by Xploreinfinity ✧･ﾟ: ✧･ﾟ: :･ﾟ✧:･ﾟ✧
