Amazon

The exercise was a high-level architecture for cart-reminder notifications on Amazon.

• Cart stateSaved carts are read periodically to find reminders.
• User preferencesPending notifications are determined from user notification preference.
• DeliveryNotifications go through a message queue to email, SMS and push.
• TrackingDelivery and user actions are tracked and written back to user state.

This grouping is a way of reading the diagram, not a claim about team ownership.

• Experience & ingress

  Takes requests from the mobile app or web interface and routes them to a front-end server.

  • User
  • Load Balancer
  • Front End App Server

• Product & cart

  Reads products, saves items to the cart and displays saved cart contents.

  • Search Product
  • Add to Cart
  • Products Database
  • Cart Database

• User state

  Holds user details, search history, cart details per user and notification preferences.

  • User Database

• Notification orchestration

  Checks saved carts periodically, records pending notifications and decides which to send based on user preference.

  • Cart monitoring service
  • Notification Database
  • Notification sending service

• Delivery

  Queues notifications, then sends them by email, SMS or push notification.

  • Messaging Queue
  • Notification Handler
  • Subscribers

• Tracking

  Tracks delivery and user actions and writes the actions back to the User Database.

  • Notification Tracking

A native reconstruction of the original diagram. Each component lists its connections using the diagram's own labels; groups run roughly in the order a cart reminder travels.

Each step is read from the original diagram.

1. 1

   User enters

   Loads the app from mobile or web; the request goes through the load balancer to a front-end server.

   Load Balancer
2. 2

   Cart state is created or read

   Products are read, search history is saved, and items added to the cart are written to the cart database.

   Cart Database
3. 3

   Carts are checked periodically

   The cart monitoring service checks saved cart contents at periodic intervals.

   Cart monitoring service
4. 4

   Pending state is recorded

   Pending notification details are updated in the notification database, alongside user notification preferences.

   Notification Database
5. 5

   Notification is queued

   The sending service determines pending notifications from user preference and adds them to the message queue.

   Messaging Queue
6. 6

   Notification is sent

   The handler reads from the queue and sends via email, SMS or push notification.

   Notification Handler
7. 7

   Tracking captures outcomes

   Notification tracking follows delivery and user actions.

   Notification Tracking
8. 8

   User state is updated

   User actions on notifications are written back to the User Database.

   User Database

The HLD does not state why it was drawn this way. The original diagram is shown first. These are separated into what it shows and what I read into its structure.

View full HLD Expand

Shown in the HLD

• Separate data stores

  Products, users, carts and notifications each have their own database.

• Periodic cart monitoring

  A cart monitoring service checks saved carts at intervals and records pending notifications.

• A message queue before delivery

  The notification sending service adds notifications to a message queue; a separate handler reads from it.

• Multi-channel delivery

  The handler sends by email, SMS or push notification.

• Separate notification tracking

  Tracking follows delivery and user actions, and user actions are written back to the User Database.

Architecture interpretation

• The queue separates generation from handling

  Placing the queue between the sending service and the handler lets the two evolve and run independently.

• Eligibility is out of the request path

  Because monitoring runs periodically in its own service, reminder eligibility is not computed while a user is browsing.

• Stores isolate state domains

  Splitting user, cart, product and notification data keeps each domain's state separate.

• Tracking closes the loop

  Writing user actions back to the User Database gives the notification path a feedback signal.