SugarCRM Amazon Fraud Detector is a NodeJS app that automates the fraud detection workflow by creating and tearing down AWS resources. This Node app uses a standard architecture that should be really easy to follow and most importantly, it takes care of all the dependencies that are required while setting up AWS.
- NodeJS v16.5.0 (or newer) installed. Check Install NodeJS for help.
This project contains few dependencies that need to be installed for it to work, those are downloaded from npm's public registry.
Follow these steps to clone and install those dependencies.
- Clone this repository
- Enter in the folder
- Install dependencies
git clone https://github.com/sugarcrm-developers/sugar-fraud-detector.git
cd sugar-fraud-detector
npm install
There are two types of configuration for this project to work, one is related to AWS, where we set them in the .env file and the other is related to the Fraud Detector resources you will create in your account in AWS.
AWS access and secret keys were provided by step five when creating a new user, you should have them, just copy them and place on those placeholders down below as well as choose a region where you’d like to have your Fraud Detection setup.
If you dont have or lost your access/secret keys, check this post, it helps you to generate a new one.
As part of this project we've provided a template called .env-template that you can rename it to .env and update with your data accordingly.
AWS_ACCESS_KEY_ID=<YOUR KEY>
AWS_SECRET_ACCESS_KEY=<YOUR SECRET>
AWS_API_REGION=<YOUR REGION>
AWS_DEFAULT_REGION=<YOUR REGION>
DEBUG=sugar:frauddetector:*
This file is located at src/config/index.js and contains the metadata for SugarFraudDetector uses to create all the necessary resources in AWS. It is self-explanatory and has been coded in a resourceful way so you can just update/adapt for your needs. Within fraudDetector you will notice there’s definitions for account and transaction, each with its own metadata as well as a variables section which contains the fields your model uses to train subsequently used in your request events when asking for predictions from AWS service.
It has been configured to perform the tasks described in the blog series, but feel free to change as needed.
There are two apps create and teardown that you can execute each will do its job respecting all the dependencies required by AWS. Both apps use the metadata configuration above to perform its tasks. Both apps provide an extensive logging capabilities so you know what the app is doing.
Reads the metadata configuration and creates all resources in AWS required by Fraud Detector. It does create IAM roles and permissions, S3 bucket, uploads files and obvioulsy Fraud detector resources.
npm run createTearing down all the resources is as important, if not more, than creating them, this is where you stop paying for your resources (if you’re not in the trial period of course).
npm run teardownAs pointed out in our Adding AWS powered fraud detection to Sugar Sell - Part 2
, this script stored in src/scripts/filter-by-date.py Python script to help filtering and preparing your data for public datasets.