Bitcoin-forex

Problem Statement:

To develop a service that constantly checks the currency exchange rate from Bitcoin to US Dollars (1 Bitcoin => X USD)

Approach:

• Research and Analysis of Bitcoin APIs and services
• Create a flask app service around the Bitcoin wrapper function
• Create 2 REST API endpoints - getLatestRate and getHistoricalRates
• Transformation of data and mapping the API response to corresponding endpoints
• Return the JSON API results
• Testing the service and writing test scenarios covering all edge cases

Architecture Diagram

Overview of "Bitcoin-forex" Workflow:

1. Create Flask API (GET - /api/getLatestRate)

   • An API/wrapper around the Bitcoin service
   • Gets Bitcoin's latest conversion rate(USD) by calling "get_latest_price()" method
   • Returns the json data with the latest USD price

2. Create Flask API (GET - /api/getHistoricalRates)

   • An API/wrapper around the Bitcoin service
   • Converts startDate and endDate to corresponding datetime format to query Bitcoin service
   • Check period is configurable
   • Gets Bitcoin's historical rates(USD) by calling "get_previous_price_list()" method between startDate and endDate
   • Returns the json data with the collection of historical rates

3. Create tests for Bitcoin service

   • Unit test cases written by using forex bitcoin service and pytest framework
   • Covered all edge cases and failure cases for both get_latest_price() and get_previous_price_list() methods

API usage:

1. Get Latest Rate (As of 5th Nov, 14:00 IST) -- https://bitcoin-forex.herokuapp.com/api/getLatestRate

   response:
   	{
   	    "latest_rate": 14460.51
   	}
   

2. Get Historical Rates (As of 5th Nov, 14:00 IST) -- https://bitcoin-forex.herokuapp.com/api/getHistoricalRates?startDate=2020-10-15&endDate=2020-10-25

   response:
   	{
   	    "historical_rates": {
   		"2020-10-15": 11510.5367,
   		"2020-10-16": 11325.5217,
   		"2020-10-17": 11369.9133,
   		"2020-10-18": 11516.9667,
   		"2020-10-19": 11758.5467,
   		"2020-10-20": 11922.975,
   		"2020-10-21": 12811.4867,
   		"2020-10-22": 12987.9017,
   		"2020-10-23": 12940.1067,
   		"2020-10-24": 13127.055,
   		"2020-10-25": 13039.0133
   	    }
   	}
   

Project Structure:

1. bitcoin.py -- An API/wrapper function for the Bitcoin service
2. tests/test_bitcoin.py -- Provides unit testing for Bitcoin service
3. architecture.png -- Architecture diagram of the project
4. Procfile -- A file which is used to run a web app deployed on Heroku
5. requirements.txt -- Dependency libraries for the project
6. runtime.txt -- A file which contains runtime version on which the app is running

Pre requisites:

• Python 3.7 or Python 3.8
• Flask

Libraries Used:

1. flask_restful -- Extension for Flask that adds support for quickly building REST APIs.
2. forex-python -- Service that provides Foreign exchange rates and currency conversion.
3. pytest -- Testing framework based on Python.
4. flask -- Web application framework written in Python which is based on Werkzeug WSGI toolkit and Jinja2 template engine.

Library Installation:

• Install virtualenv package to create a virtual env --> pip3 install virtualenv
• Create virtual env to install the dependencies --> virtualenv .env
• Activate the virtual env --> source .env/Scripts/activate
• Install the dependency libraries (Provided in requirements.txt file) --> pip3 install -r requirements.txt

Deployment options:

Hooray! Now our Bitcoin service is ready to use. To make this application production ready, we have several options:

1. Deploying on Heroku

   • Create a Procfile which is used to run a web app deployed on Heroku
   • Create requirements.txt file which is used as dependency libraries for the project
   • Create an app on Heroku and connect to the GitHub where the project is pushed
   • Create a deployment pipeline in Heroku

2. Deploying on AWS

   • Create API gateway REST APIs for both the endpoints
   • Additionally, API keys can be created for the REST APIs to be secured
   • Create a lambda function which corresponds to API gateway (trigger to lambda function)
   • Create a CI/CD pipeline using code build, code deploy to automatically trigger the build and deploy the service

AWS resources that can be used to deploy the service:

• API gateway -- To create REST APIs
• AWS Lambda function -- To write code without having to worry about infrastructure and scale automatically
• SSM parameter store -- To secretly store the credentials, URLs and API keys
• AWS Code Pipeline -- To automate release pipelines
• AWS CodeBuild -- To compile source code, runs tests, and produces software packages that are ready to deploy
• AWS CodeDeploy -- To deploy service that automates application deployments to Amazon EC2 instances, serverless Lambda functions, or Amazon ECS services.
• AWS KMS -- Optionally we can use AWS Key Management Service to encrypt data and to automatically rotate customer master keys(CMK)
• AWS S3 -- Optionally we can use Simple Storage Service to store metadata information and to host any static web pages, sites.
• AWS DynamoDB -- Optionally we can use DynamoDB database to store items for high availability and durability and to offload the administrative burden

Solution:

This service is deployed on Heroku and always gets the latest Bitcoin conversion rates (USD) on the date the API is triggered. Find the below API endpoints of the service.

• To get latest rate --> https://bitcoin-forex.herokuapp.com/api/getLatestRate

• To get historical rates --> https://bitcoin-forex.herokuapp.com/api/getHistoricalRates?startDate=2020-10-15&endDate=2020-10-25

PS: Bitcoin service demo screenshots available here