Concepts, technologies, patterns and quick references for ML interviews and system design. Each topic gives you the short version first, goes deeper when you want it, and links to the Lessons and Problems that use it.
How to split long documents or event streams into pieces a model can use, the main ways to choose boundaries, and what each choice costs in recall, context use and complexity.
How to put an interval on a pass rate, compare two versions scored on the same cases with a paired analysis, and plan how many cases a comparison needs. Covers the assumptions that make a measured gain misleading.
Run lexical (BM25) and vector retrieval on the same query and merge the two ranked lists with reciprocal rank fusion. Covers what the merge gains and costs, where it fails, and how to test it on your own corpus and queries.
What a model's key-value cache holds during generation, how its memory grows with model shape and context length, and how that limits concurrent requests. Separates it from prompt (prefix) caching and from application-level response caching.
A reusable way to structure answers to behavioral interview questions: Situation, Task, Action and Result, with reflection where it helps. Covers when the structure fits, where it goes wrong and how to keep answers natural.
Topics: behavioral-interviews · star-method · communication