Case Studies for System Design Interviews

Caching geofences handles millions of SDK requests by serving data from the cache, boosting performance and reducing backend traffic for the geofencing service.

Explore how a webhook delivery pipeline uses an events service, a message queue, and a deliver service to retry failed requests with exponential backoff, decoupled queues, and dead-letter handling.

Explore a messaging solution that supports one-to-one and group messaging, presence, read and delivery status, file sharing, location sharing, and audio/video calls across mobile, desktop, and web clients.

Explore a backend messaging design where an api routes client requests to messaging, delivery, and acknowledgement services, stores read/delivery status in a messaging database, and supports conversations, contacts, and groups.

Explore time-range sharding for message systems by partitioning data monthly. Analyze read/write traffic distribution and capacity planning for current-month bottlenecks when events alter data volume.

Learn hashing-based sharding using a conversation id as the partition key and modulo-n distribution for data. Ensure that single-conversation reads and writes stay on a single shard, reducing cross-shard queries.

Explore how replication enhances fault tolerance in the message history database, preventing data loss by using primary/secondary or multi-master setups and promoting a secondary to primary during failures.

Store message history on the client to scale for millions of users and enable local or cloud backups, while the backend stores messages temporarily and transfers them between users.

Explore the design of a mobile push notification service, defining user segments by activity, location, or VIP attributes, monitoring statuses like sent, clicked, failed, and routing to push providers.

Explore how a push notification system orchestrates backend, operators, SDKs, and providers to define segments, send targeted messages, and track delivery and engagement.

Define an operator-friendly high level design that routes backend requests via an api service to applications, segmentation, delivery, and notification services.

Explore how segmentation data grows with per-device custom attributes and location history, and calculate storage needs for 100 million devices.

Analyze segmentation writes for static device properties like the device operating system, app version, and country; contrast dynamic last active time and location history with small per-request data.

Operators define segmentation criteria with multiple conditions, store location data in a separate database via vertical partitioning to improve indexing and scalability, while the application layer becomes more complex.

Apply rate limiting to protect system availability by enforcing per-user API limits and using tokens or API keys to identify users, returning 429 when exceeded.

Learn how a decoupled notification and delivery system uses asynchronous messaging to meet high throughput and availability. Multiple service instances publish to a queue, enabling parallel processing and fault tolerance.