Requirements, metrics, architecture and tradeoffs — the framework behind every design interview and every real system.
10h · 8 lessons · 1 challenge
Before this: Data Pipelines & Feature Engineering, Structuring ML Projects
A repeatable framework: clarify the requirements, define online and offline metrics, size the problem, sketch the architecture, then name the tradeoffs before anyone asks.
Worked designs include a feed ranker, a fraud detector, a search system and an LLM assistant, each done end to end. This module and the interview track are the two most cited by learners who move up a level.
You present an architecture and the game pushes back the way an interviewer does — traffic times ten, latency halved, budget cut. Adapt or justify.
Worth 700 XP.
