To run everything needed for the application, run this in the root of the repository

docker-compose up -d

This builds the API service, the notifications service and runs Postgres and Kafka. To be able to send emails edit docker-compose and set these environment variables with eg. Mailtrap plain credentials

SMTP_HOST: enterHost
SMTP_PORT: 587
SMTP_USERNAME: smtppw
SMTP_PASSWORD: smtpu

API service provides the following endpoints:

POST /register

{
  email: string,
  password: string
}

GET /verify/:email/:code

POST /login

{
  email: string,
  password: string
}

GET /protected

• Email verification using notification service with notification sent over Kafka. After registration another email notification is sent- Welcome email
• Auth using JWT and a test endpoint /protected that is protected route and checks JWT provided in the header

Notification service accepts messages over a Kafka topic (any MQ or Pubsub can be implemented) and in a consumer group processes these notifications. Currently Email notification is implemented and SMS notifications mocked. Any new notification type processor can be implemented.

• Simple SMTP u/p based auth email service
• Compilation of HTML templates that are sent in the mail
• Horizontally scalable with the Kafka consumer group

• Unit tests are not written at this time, but the code is written with tests in mind and everything should be easily testable.
• Notifications are currently written as one notification type per Kafka message. Ability to send the same notification over multiple processors can be added. For instance: -> User preferences set to receive notifications over both Email and SMS. -> Produce a notification (eg. Welcome message with additional metadata in the message being First name and Last name) -> Notification is sent as only one Kafka message alongside user preferences -> Both SMS and email is sent
• While notifications being in another service brings operational complexity and need for orchestration of the services. On the other hand it brings us the ability to separately scale/deploy the service and persist the notification messages with using Kafka