Design a RAG-powered LLM application
A complete 45-minute mock covering retrieval, cost/latency tradeoffs, and quality guardrails.
第一個成果
You can quote the dollars-per-query cost of your design.
適合對象
Engineers prepping for FAANG and AI-native system design interviews
需要
Mac, Microphone, Whiteboard tool (Excalidraw, Figma, etc.)
範例成果
You can quote the dollars-per-query cost of your design.
A complete 45-minute mock covering retrieval, cost/latency tradeoffs, and quality guardrails.
模擬面試
階段完成
6
需求、核心實體、API、整體設計、深入主題與總結，全部用具體數字走完一輪。
深入主題
兩到三個子系統
你主動帶起最具張力的取捨，必要時搭配 sub-100ms 的具體數字。
總結
清楚說出這個設計優化了什麼、捨棄了什麼。