APMS is a code name, it helps businesses manage their marketing efforts. It allows businesses to track the performance of their affiliates, manage ads, payments and marketing strategies.

Proof of Skills used in this project:

• .NET 8
• JWT Authentication
• Mediatr
• Mapster for Mapping
• Fluent Validation
• EF Core
• Postman Collections for tests and examples

• Role Based Authorization
• JWT Authentication
• Clean Architecture
• CQRS: Command Query Responsibility Segregation
• API Versioning
• Unity of Work
• Tests
• Docker

To use the APMS Application, you need to have Docker installed on your machine.

Run the following command to start the application:

docker-compose up

docker-compose.yml

webapp: This container runs the main application, which is built using .NET 8. It communicates with the db container to perform database operations. port:8000

db: This container runs SQL Server, which is used to store and manage the application's data. It receives requests from the webapp container and performs the necessary database operations.

entrypoint.sh will create the database and add data for tests

Script for reference: DataBase Seeder

This layer is responsible for handling HTTP requests and responses. It includes routing, serialization, error handling, and other web-related tasks. It communicates with the Application layer to process requests.

• AdminUser:
  • login: admin@example
  • password: password-test
• _AffiliateUser:
  • login: admin@example
  • password: password-test

It processes requests from the WebAPI layer, performs necessary computations or transformations, and communicates with the Domain layer for data.

CQRS - command query responsibility segregation

• The Command stack is responsible for handling all commands, which are requests that modify the state of the system.

• The Query stack is responsible for handling all queries. It is great when we need to access differente data strucutures, like: consolidated data or real time analysis.

This separation allows for flexibility in scaling, as the Command stack and Query stack can be scaled independently based on their load. It also improves performance, as read and write operations can be optimized separately.

The Domain layer represents the business objects of the application. It includes entities, value objects, and aggregates. This layer is isolated and should not depend on any other layers.

The Infra layer provides concrete implementations of the interfaces defined in the Domain layer including persistence.

Also, the dependency injection is there.

[ApiController]
[Route("api/admin/[controller]")]
[Authorize(Role.Admin)]
public class AdminAffiliateController(IMediator mediator)...

[Route("api/[controller]")]
[ApiController]
[Authorize(Roles = Role.Affiliate)]
public class AffiliateController(IMediator mediator) ...

http://localhost:5043/api/admin/AdminAffiliate/[affiliateId]/customers

   [
    {
        "id": 1,
        "email": "[email protected]",
        "fullName": "John Doe",
        "genderId": 1,
        "genderName": "Female",
        "birthDate": "1990-01-01",
        "avgTicket": 600.00,
        "totalPurchase": 10000.00
    },
    {
        "id": 2,
        "email": "[email protected]",
        "fullName": "Jane Doe",
        "genderId": 2,
        "genderName": "Other",
        "birthDate": "1992-02-02",
        "avgTicket": 1000.00,
        "totalPurchase": 200000.00
    }
]

You can use the Postman to perform API tests.

1. Download the APMS Postman Collection;

2. Prepare the environemtn

3. Click on "Run Collection"

Flows and Tests are written in Javascript.

{
  "FullName": "{{$randomEmail}}",
  "email": "{{$randomEmail}}",
  "password": "{{password}}"
}

pm.test("Status code is 2xx", function () {...});
...
var affiliateId = pm.response.json().data.id;
pm.environment.set("affiliateId", affiliateId);

Added the most important unit tests based on affiliate and customer scenarios.

Mock example of data from Redis

[
    {
        "adId": 1,
        "endTime": "2024-01-16T01:47:23.5372804-03:00",
        "productName": "Product 1",
        "productId": 1,
        "stats": {
            "totalViewed": 1,
            "totalPurchases": 1,
            "totalPurchasesAmount": 1,
            "visualizationsLastDay": 1,
            "purchasesLastDay": 1
        }
    },
    {
        "adId": 2,
        "endTime": "2024-01-17T01:47:23.5372848-03:00",
        "productName": "Product 2",
        "productId": 2,
        "stats": {
            "totalViewed": 2,
            "totalPurchases": 2,
            "totalPurchasesAmount": 2,
            "visualizationsLastDay": 2,
            "purchasesLastDay": 2
        }
    },

Redis can be used for real time stats:

flowchart LR
    AdViewedEvent[/AdViewedEvent/] --> AdEventTracker
    AdPurchasedEvent[/AdPurchasedEvent/] --> AdEventTracker
    AdEventTracker -->|Sends events| Kafka[/Kafka/]
        Kafka -->|Streams events| AdEventProcessor[AdEventProcessor]
        Flink -->|Processes events| OLAP[(OLAP)]
    AdEventProcessor -->|Updates counters| Redis[(Redis)]
    Kafka -->|Streams events| Flink[Apache Flink]

Please, click on the links bellow to see the full documentation of the project.