Core

Chunking

Chunking splits documents into the passages you embed and retrieve, and it is one of the highest-leverage knobs in RAG. Too large and embeddings are diluted so retrieval is imprecise; too small and chunks lose the context needed to answer. Beyond fixed-size splitting, structure-aware and semantic chunking keep coherent units intact, and parent-child (small-to-big) retrieval matches on small chunks but returns larger context. Applied-AI interviews probe it because poor chunking silently caps retrieval quality.

a free account unlocks the core curriculum tier · no card

RELATED CONCEPTS

The RAG Pipeline Embeddings Vector Search and ANN Indexes Hierarchical Retrieval (RAPTOR and Small-to-Big)

PRACTICE THIS IN REAL QUESTIONS

RAG & Agent System DesignCompare chunking strategies (fixed-size, recursive, semantic, parent-child). How do you pick?→RAG & Agent System DesignWhat is semantic chunking, and how does it compare to fixed-size chunking?→RAG & Agent System DesignWhat is parent-child (small-to-big) retrieval, and why does it improve RAG?→Coding & DSAWrite three chunkers (fixed-size with overlap, recursive separator, semantic) and defend when each wins.→RAG & Agent System DesignWhat is HyDE (Hypothetical Document Embeddings), and why does it improve retrieval?→RAG & Agent System DesignWhat is contextual retrieval, and how does it fix the lost-context problem in chunking?→

COMPANIES THAT ASSUME THIS

OpenAI Anthropic Google Microsoft Scale AI Cognition

NEXT IN RETRIEVAL & AGENTSReranking→