Watch a demo of our Chatroom in action
- React-based component
- Markdown, Images, Buttons
- Customizable with SASS variables
- Generates a unique session id and keeps it in
sessionStorage - Queues consecutive bot messages for better readability
- Speech input (only in Chrome for now)
- Demo mode included (ideal for scripted screencasts)
- Hosted on S3 for easy use
- Includes a
BotServerChannelfor use with Rasa Core (underrasa_utils)
Embed the chatroom.js in the HTML of your website and configure it to connect to your Rasa bot. (see below)
<head>
<link rel="stylesheet" href="https://npm-scalableminds.s3.eu-central-1.amazonaws.com/@scalableminds/chatroom@master/dist/Chatroom.css" />
</head>
<body>
<div class="chat-container"></div>
<script src="https://npm-scalableminds.s3.eu-central-1.amazonaws.com/@scalableminds/chatroom@master/dist/Chatroom.js"/></script>
<script type="text/javascript">
var chatroom = window.Chatroom({
host: "http://localhost:5005",
title: "Chat with Mike",
container: document.querySelector(".chat-container"),
welcomeMessage: "Hi, I am Mike. How may I help you?",
speechRecognition: "en-US"
});
chatroom.openChat();
</script>
</body>Note, the version of the Chatroom's Javascript file is encoded in the URL. chatroom@master is always the latest version from the GitHub master branch. Use e.g. chatroom@0.8.2 to load a specific release. All Releases can be found here.
| Chatroom Version | Compatible Rasa Core Version |
|---|---|
| 0.9.x | 0.11.4+ |
| 0.8.x | 0.11.4+ |
| 0.7.8 | 0.10.4+ |
- Clone repository
- Install frontend dependencies
yarn install - Build frontend files
yarn build - Create an HTML page for your Chatroom (see usage example above or modify index.html)
- Make sure to adjust the
hostoption. Usehttp://localhost:5005when testing locally - Run
yarn serveto launch a web server with your Chatroom onhttp://localhost:8080 - Integrate with a Rasa Core project (see standard or custom project below)
- Copy
rasa_utilsto your project - Install the Python dependencies from
rasa_utils/requirements.txt - Run your bot with
python -m rasa_utils.bot -d models/current/dialogue -u models/current/nlu - The bot server will be available at
0.0.0.0:5005
- Copy the
rasa_utils/bot_server_channel.pyto your project - Install the Python dependencies from
rasa_utils/requirements.txt - Create or Modifiy
credentials.ymlfile and add:
rasa_utils.bot_server_channel.BotServerInputChannel:
# pass
- Start the Rasa bot using the command line and pass the
--credentialsflag:
python -m rasa_core.run \
--core models/current/dialogue \
--nlu models/current/nlu \
--endpoints endpoints.yml \
--credentials credentials.yml
- Copy the
rasa_utils/bot_server_channel.pyto your project - Install the Python dependencies from
rasa_utils/requirements.txt - Register the
BotServerInputChannelwith your Rasa CoreAgent(see below) - Run your bot. By default the bot server will be available at
0.0.0.0:5005
from bot_server_channel import BotServerInputChannel
# Creating the Interpreter and Agent
def load_agent(): ...
# Creating the server
def main_server():
agent = load_agent()
channel = BotServerInputChannel(agent, port=5005)
agent.handle_channels([channel], http_port=5005)
main_server()yarn install
yarn watch
yarn serve
Open http://localhost:8080/demo.html in your browser.
yarn build
Distributable files will be created in folder dist.
AGPL v3
Made by scalable minds