What software architecture is, why it decides how a system ages, and what the architect actually does. You learn to read stakeholder concerns, constraints, and quality attributes as the forces behind every structural decision.

Coupling, cohesion, and the design principles that keep large systems changeable. You practice drawing module boundaries and use domain-driven design to make those boundaries match the business.

The major monolithic and distributed architecture styles, from layered and modular monoliths to microservices, event-driven, and serverless. You learn what each style optimizes for, and how to choose one and migrate without breaking a running system.

The rules that change when a system spans machines: network fallacies, CAP, and consistency models. You design service communication, then replication, sharding, sagas, and CQRS for data that no longer fits one database.

How systems stay fast while load grows tenfold: stateless design, load balancing, caching at every layer, and queues that absorb spikes. You then learn to measure latency honestly, find real bottlenecks, and plan capacity.

Why every dependency fails eventually, and how to keep serving users anyway. You quantify availability with SLOs and error budgets, design redundancy and failover, and apply resilience patterns like circuit breakers, backed by observability.

Security as a structural property you design in, not a feature you add later. You model threats with STRIDE, architect identity with OAuth 2.0 and OpenID Connect, and protect data with encryption, secrets management, and hardened APIs.

The daily craft that separates working architects from diagram owners. You document with C4 views and ADRs, evaluate designs with scenarios and fitness functions, manage technical debt, and modernize legacy systems without a risky rewrite.