The goal of this technical test is to check if the candidate can build quality software to solve business problems.

Disclaimer: this is a toy problem not an actual project :-)

Artists who publish their music on our music streaming platform want to know and reach to their fans. We need to provide tools to help them connect to their fan base.

The tools to develop:

1. Fanbase: a web API that provide aggregated information about the fanbase of an artist.
2. TopFans: a script that displays the artist top fans information, so the artist can send them VIP concert tickets and other goodies.

For legal reasons:

1. Users who do not consent to tracking are excluded from the result produced by Fanbase and TopFans tools. See the consent field in the database.
2. Minor (underage) users information is not displayed by TopFans.

• All the programs (Fanbase web API & TopFans script) should run inside Docker containers.
• The containers should be organised with a Docker compose configuration.
• A MariaDB database container is provided with the user data. See database/ directory.

• Fanbase should be built as follows:

  docker-compose build fanbase

• Then, the web API should be started with the following command:

  docker-compose up fanbase

• HTTP requests should be sent the Fanbase web API like the example bellow:

  curl -XPOST https://localhost:8080/fanbase -d '{"artist_id": 123}'

• The expected response is a JSON object with the following fields:

  Field	Type	Description
  success	bool	the artist have at least one fan identified fan
  artist_id	int	ID of the artist, provided by the HTTP request
  total_fans	int	size of the artist fan base
  fans_median_age	int	median age of the artist fans
  fans_top_countries	array[string]	top 3 countries with the most fans (ordered, top country first)

  Example result:

  {
    "success": true,
    "artist_id": 123,
    "total_fans": 456,
    "fans_median_age": 21,
    "fans_top_countries": ["TN", "FR", "NE"]
  }

• TopFans script is partially written in the topfans/run.py file. But this version don't work and should be refactored.

• The refactored and working script should be written in Python.

• The script should be set-up with this command:

  docker-compose build topfans

• Then it should be executed as follows:

  docker-compose run topfans --artist_id 123 --nb_fans 3

• The expected output should be a CSV content with the following fields:

  Field	Type	Description
  rank	int	rank of the fan, the top fan have the rank = 1
  artist_id	int	ID of the artist, provided as input parameter
  user_id	int	ID of the user
  country	string	user's country
  score	float	from 0 = not fan of the artist to 1 = die-hard fan

  Example result:

  rank,artist_id,user_id,country,score
  1,123,987,"TN",0.91
  2,123,654,"FR",0.90
  3,123,321,"TN",0.74
  

1. A working, maintainable and easy to read source codes. All the project files should be compressed into a single zip file.
2. Including a NOTES.md file where you explain your design and implementation choices.

Not required but nice to have:

• Python coding style compatible with the PEP8.
• Unit tests can be added with notes about how to run them.
• Authentication support can be added to Fanbase web API.
• Improvements to the tools can be proposed in the NOTES.md file.

If you have any issue or any question, please contact us and we'll help you.

Happy coding.