Charlie the Chatterbot is a Machine Learning Chatbot that integrates with Slack. Place the Slack bot into your channel and the bot will be automatically trained on all of the conversations in that particular channel. Over time, the bot will learn and represent the culture of the Slack channel. All of the data that the bot learns from is also analyzed through Alchemy's sentiment analysis, and displayed in graph form.

We are live at uai.website!

• Installation
• Configuration
• Architecture
• Contributing
• Authors

If you need to set up a Python dev environment, check out this guide in our wiki.

From the root directory,

npm install
pip install -r requirememnts.txt

This project uses mysql for the node/express server, and Mongodb for the chat bot. Both must be installed and running.

$ mysql -u root < server/Schema.sql

If your MySQL instance needs a password, add -p to this command

$ mongod

It is recommended that each of the following commands is run in it's own terminal window. This will help with development flow as you will be able to watch webpack, the server and the bot at the same time.

$ npm run webpack

$ nodemon server/server.js

$ nodemon server/chatterbot/chatterbot_entry.js

The following files must be added to the server/config folder. These files hold secrets, such as API keys.

AlchemyApiKey.js

module.exports = <ALCHEMY_API_KEY>;

botKey.js

module.exports = <SLACK_BOT_KEY>;

pythonSettings.js

module.exports = {
  mode: 'text',
  pythonPath: '<PATH_TO_PYTHON>',
  args: [0, 'test', 0],
};

winston.js set logging levels - uncomment lines for loggly integration.

const winston = require('winston');
require('winston-loggly');

module.exports = (logger) => {
  logger.level = 'info';
  // logger.add(winston.transports.Loggly, {
  //   token: '<LOGGLY_API_TOKEN>',
  //   subdomain: 'your loggly subdomain',
  //   tags: ['Winston-NodeJS'],
  //   json: true,
  // });
  // logger.add(require('winston-daily-rotate-file'), {
  //   filename: 'server.log',
  //   dirname: './logs',
  //   prepend: true
  // });
};

• React
• D3.js
• Webpack, see this

• AlchemyAPI - Sentiment analysis
• Slack RTM

• Node/Express
• Chatterbot
• Socket.io
• MongoDB
• MySQL

• Travis CI - Continuous Integration Testing
• Mocha/Chai
• Karma

Our deployment strategy is explained in more detail in our wiki.

See CONTRIBUTING.md to see the Git Flow that we used in this project.

• Jin Bok - Product Owner
• Mark Tyneway - Scrum Master
• Jovani Marinez Rico - Developer Specialist
• Josh Wentworth - Developer Specialist

MIT