Anti Fraud Service

What Is It?

According to the article Fraud Detection With GPS and Airplanes, It's possible to calculate the distance between two given cities and make sure it’s realistic to travel the distance in the given time via airplane.

For example: If Alice makes a transaction in Berlin DE at 05 Aug 12 02:47 UTC and 10 hours later she makes one in Toronto CA at 05 Aug 12 14:30 UTC we cannot flag this transaction as fraudulent since it is mathematically possible.

However, if there is an attempt to do a transaction in Cancun MX at 05 Aug 12 15:03 UTC, this time it would be flagged as fraudulent. This is because is not possible for Alice to travel from Toronto to Cancun in that amount of time.

How It Works?

How To Run

1. Clone the project
2. Run glide install
3. Set up an .env file in the project's root directory

POSTGRES_DB=anti-fraud-api
POSTGRES_PASSWORD=mysecretdbpassword
POSTGRES_USER=postgres
DB_TYPE=postgres
DB_HOST=localhost
DB_HOST_DOCKER=host.docker.internal
DB_PORT=5432
CACHE_PORT=6379
CACHE_PASSWORD=mysecretcachepassword
PORT=8000

4. Simulate an environment with Redis (cache), Postgres (db), and the Application by running docker-compose up --build

How To Use

You can use postman or curl in order to post a new transaction in the following endpoint: localhost:8000/new

First Transaction

This transaction should be a legit one since it is the first one and it can happen anywhere in the world.

Sample input

{
	  "user_id" : 2,
	  "city_name" : "Berlin",
	  "country_code" : "de",
	  "time" : "05 Aug 12 02:47 UTC"
}

Sample Output:

{"message":"success","status":true,"tx":{"ID":1,"CreatedAt":"2019-06-17T02:54:11.6864658Z","UpdatedAt":"2019-06-17T02:54:11.6864658Z","DeletedAt":null,"city_name":"berlin","country_code":"de","time":"05 Aug 12 02:47 UTC","user_id":1}}

Second Transaction

This one should also be legit. It is possible to travel from Berlin to Toronto within the time difference between the previous transaction (Berlin) and the current transaction (Toronto).

Sample input

{
	  "user_id" : 2,
	  "city_name" : "toronto",
	  "country_code" : "ca",
	  "time" : "05 Aug 12 14:30 UTC"
}

Sample Output:

{"message":"success","status":true,"tx":{"ID":2,"CreatedAt":"2019-06-17T04:17:18.5402872Z","UpdatedAt":"2019-06-17T04:17:18.5402872Z","DeletedAt":null,"city_name":"toronto","country_code":"ca","time":"05 Aug 12 13:30 UTC","user_id":2}}

Third Transaction

This one should be flagged as fraud since it is impossible to travel from Toronto CA to Cancun MX in just 33 minutes.

Sample input

{
	  "user_id" : 2,
	  "city_name" : "cancun",
	  "country_code" : "mx",
	  "time" : "05 Aug 12 15:03 UTC"

}

Sample Output:

{"message":"Transaction is Fraudalent","status":false}

References

[0] https://inversegravity.net/2019/fraud-detection-with-GPS-and-airplanes/