helxm/artificial-trader

Crypto-currency trading engine written in NodeJS

Artificial Trader

Milestones

Setup

• Set up a Node server and deploy it to Heroku
• Setup a Mongo instance in mLab and connect to it from Node
• Deploy Client app to Surge.sh and point server calls to server's APIs

Server

• Connect to GDAX sandboxes for development and production environment
• Handle security based on GDAX/Coinbase account
• Listen to changes on crypto-currencies
• Hold feeds in a local data structure
• Fetch user account info from GDAX
• Define Rule class
• Fetch user rules from database
• Enable Buy/Sell, Taker/Maker actions based
• Perform transactions based on rules
• Store transactions in database
• Expose a REST API to fetch user information, performance and any stored transaction(mode to come here...)
• Expose a WebSocket API to fetch feeds on both user performance and GDAX feeds
• Alert via email/SMS/native notifications when a transaction happened

Client

• Enable new users to enter their Coinbase credentials
• Enable user to create or reuse rules. Which will affect the percentage-risk for automated trading
• Enable user to start/stop their artificial traders
• Connect via Websockets to show transactions realtime
• Show history and performance

Docs

3rd Party APIs

• Stocks feed: IEX Trading
• Stocks order management(in the near future...): Robinhood's Private API
• Crypto-currencies feed and order management: GDAX Websocket Client

Rules

I define a Rule as a single instance of multiple trading strategies to be used by the Engine. Rules can be defined by the user and will be stored in the Mongo instance ######Attributes // TODO

Additional thoughts:

• Create both Desktop and Web versions and share the code as much as possible
• Enable the user to select what currency to watch
• Enable user settings in general including the above mentioned and also rules created and current rules used
• Enable notifications by email or text messages. Also use the native notifications API in the browser