System Design Interview: Architectural Thinking in Big Tech

I've signed a deal with Piter to publish my book, System Design Interview: Architectural Thinking in Big Tech. It has 35 chapters on breaking down a problem, asking questions, spotting constraints, and explaining architectural decisions. What do you do when the requirements are incomplete, load keeps growing, and reliability and speed come at a cost?

At the heart of the book is my seven-step system design interview framework, which we introduced and scaled at T-Bank. It takes you from the problem statement and clarifying requirements through choosing technologies, operating the system, and evolving it. There's also a view from the other side of the table: what an interviewer evaluates, where candidates lose the thread, and how to figure out what to practice.

Most of the practical examples deal with AI and ML: model serving, RAG, enterprise AI assistants, coding agents, fraud detection, and recommendations. They cover concrete questions: where to check access, what to do when a model fails, how to evaluate quality, and what a response costs.

There's also a chapter with a mock interview on a RAG platform. It follows the conversation between candidate and interviewer, the clarifications and mistakes, and the diagram as it evolves. Each chapter ends with a map, a checklist, and an exercise to try reasoning through it yourself.

The book should be useful to people preparing for interviews, people conducting them, and engineers who need to justify decisions in their own systems. You can go through the full process or pick one problem and try out your approach.

I'll get 100 author's copies. The book will only be ready after I move to London, though :)

The attached image maps all 35 chapters. You can already see what's inside and pick where you'd like to start.

#Books #SystemDesign #Architecture #AI #Engineering