study8677/awesome-architecture

Awesome Architecture is a free, bilingual (Chinese and English) knowledge library that teaches you how to think like a software architect. Instead of focusing on how to write code, it helps you figure out what a system should look like before you build it, how data flows, where bottlenecks will appear, and what trade-offs you're making. The core idea is that as AI increasingly handles writing code, the truly valuable skill is knowing how to design the system itself. The repository is organized into three main sections. First, a tutorial series walks through architecture thinking step by step, from breaking down requirements to making decisions about performance, consistency, and cost, including chapters on designing AI-native systems and working with AI coding tools. Second, a templates section provides "architecture maps" for 26 real-world system types (e-commerce, social feeds, RAG knowledge bases, coding agents, etc.), explaining what problems each solves, how components fit together, and where things break at scale. Third, a cases section walks through 6 full projects from zero to production, showing not just the final architecture but the reasoning that got there. This resource is built for a range of people. A beginner learning system design can follow the tutorials in order. A founder or PM planning a new product can find the closest template and use it as a starting point. Someone preparing for a system design interview can use the templates to review common patterns. And an experienced engineer can jump straight to the "key decisions and trade-offs" sections of each template. What makes this project notable is its philosophy. It deliberately avoids discussing specific programming languages or frameworks, focusing purely on architecture decisions that transfer across technologies. It also treats architecture as something that evolves, each template includes an "evolution path" showing when a system should upgrade and when upgrading would be over-engineering. The later tutorials even address how to write architectural constraints that AI coding tools can follow, and how to review what AI produces for production-readiness.