Complete Business / Trading Bot Web App (React, Redux, Typescript, Django, MySQL, AWS, Docker, Terraform, Ansible, Nginx)

23 Dec 2021

Sign-up Flow

Login / App Flow

Please note: The account balance shown in the video is not of real money; a practice account balance (virtual money) is shown instead.

This project shows the culmination of my Full Stack and cloud skills. This project showcases a faux company “AutoEdge” which provides a web and cloud-based product: a Forex trading bot. This business never officially launched because 1. the bot was not consistently profitable (although the bot was tested to confirm that the correct implementation of the intended strategies were implemented) and 2. I wasn’t too interested in starting a trading bot business. I’m not really interested in trading in general, just was looking for an interesting project idea with a lot of meat to it.

Because the business was never launched, the payment processor was never switched to use real credit cards / PayPal (it used a practice sandbox instead with test credit card numbers). Everything else was built to completion and ready to support many users and even ready to trade with real money.

Trading Bot (Python)

• Developed a python trading bot that combined several trading strategies to trade on the Forex market.
• Scanned 28 of the major Forex pairs for trading opportunities.
• The trading bot always calculated a “stop loss” or “take profit” target made before entering into the trade.
• Ability to adjust the amount of risk to apply per trade, per currency, per day.
• Identified trend reversals, was cognizant of news that would potentially affect the market, and looked for other properties that would give the trade an edge.

Web App and Site (React / Redux / TypeScript / Django / AWS / Ansible / Terraform)

• Secure user registration requiring email confirmation.
• Seamless integration of account with OANDA Forex broker (leveraging OANDA’s API).
• Leveraged Braintree (PayPal) payment processor to handle credit card and PayPal transactions.
• Instantiated an EC2 instance, downloaded and launched Docker container (via Ansible) to start bot. Stopping the bot would stop the AWS EC2 instance. This was written to work and scale for any number of simultaneous requests / users.
• History of positions and automatic trades described via integration and use of third-party trading charts.
• History of orders and potential trades (trades the bot was considering, but ultimately did not issue a trade order) available for review.
• Account management (start and stop the bot, change country, email, username and password, manage subscription, manage link to OANDA broker). Email change requires email verification.
• SSL Certificate signed by LetsEncrypt.org

Nginx was used to route HTTP traffic to the React / Redux “static” pages while routing the HTTPS traffic to the Django APIs. Django was not used to render any of the frontend-facing pages, Django was solely used to provide a REST API.