Product-Minded System Design and Observability Roadmap

Design systems that scale and can be understood in production, with a habit of tying choices to product needs.

10 milestones · ~132 hours total

1. 1

   Requirements and estimation

   ~8h

   Clarify functional and non-functional needs, estimate traffic and storage with simple back-of-envelope maths.

   Practice this →
2. 2

   Networking and API design

   ~12h

   HTTP, DNS, load balancers, REST vs gRPC vs GraphQL, idempotency and versioning.

   Practice this →
3. 3

   Storage choices

   ~16h

   SQL vs NoSQL, indexing, replication, partitioning and picking a database based on access patterns.

   Practice this →
4. 4

   Caching and CDNs

   ~10h

   Cache-aside, write-through, eviction, invalidation problems and what to put behind a CDN.

   Practice this →
5. 5

   Asynchronous systems

   ~14h

   Queues, pub/sub, event-driven design, retries, dead-letter queues and at-least-once delivery.

   Practice this →
6. 6

   Scalability and consistency trade-offs

   ~14h

   Horizontal scaling, sharding, the CAP theorem, eventual consistency and rate limiting.

   Practice this →
7. 7

   Classic design exercises

   ~24h

   URL shortener, news feed, chat service and notification system, written up with diagrams and trade-offs.

   Practice this →
8. 8

   Observability: logs, metrics and traces

   ~14h

   The three signals, OpenTelemetry, useful dashboards and latency percentiles.

   Practice this →
9. 9

   Reliability and cost awareness

   ~10h

   Failure modes, graceful degradation, SLOs and simple cost estimates for the design choices you make.

   Practice this →
10. 10

    Product thinking in design

    ~10h

    Link design decisions to user impact, choose what to build first, write a short design doc and defend your trade-offs.

    Practice this →