The Roadmap

Zero assumptions. What a client, a server, an API, and a database actually are, where data lives, the vocabulary every design discussion uses, and how one app grows from 1 user to a million.

Learn the loop. Never memorize an answer again.

The vocabulary every LLD answer is graded on.

The 22 GoF patterns that show up in LLD interviews — each tied to the problem it solves, with Java.

The repeatable method behind every LLD answer.

18 problems, from Parking Lot to LRU Cache and Pub/Sub — with production-style Java.

Threads, locks, pools and the four shared-state bugs every LLD hides.

HTTP, HTTPS, TCP/UDP, WebSockets, REST, gRPC — the layer HLD answers sit on.

Designing APIs other teams love to use, and knowing who is calling them: resource design, pagination, versioning, idempotency and errors, REST vs GraphQL vs gRPC, authentication (sessions, JWT, OAuth 2.0, OIDC, SSO), and authorization (RBAC, ABAC, ReBAC).

Tables, ACID, isolation levels, indexes, joins, normalization.

Replication, partitioning, sharding, consistent hashing — from one box to many.

Vertical, horizontal, stateless, load balancers, CDNs, and the numbers behind it all.

Cache placement, cache invalidation, Redis, Kafka — where latency and decoupling actually live.

Boundaries, discovery, resilience, sagas and the patterns that keep a fleet from eating itself.

39 real systems across 6 levels, from URL shortener to a distributed message queue — every one answers the same 13-section template.

CAP, PACELC, consistency models, quorums, consensus, event sourcing, CQRS.

The internals senior engineers reach for: storage engines (B-tree vs LSM), unique ID generation, Bloom filters and sketches, geospatial indexes, clocks and ordering, gossip and Merkle trees, stream vs batch processing, and backpressure.

Reliability, disaster recovery, observability, security, multi-tenancy, cost — the engineer's layer.

The two templates that turn 50+ problems into one repeatable process.