Retrieval basics: chunking, embeddings and vector search
Comes up in: AI engineering, AI system design
This concept is being written. It gets the full five steps — explanation, predict, interview angle, practice with tests, and review cards. Here is exactly what it will cover, so you can study it elsewhere in the meantime.
What it covers
- RAG idea (Lewis et al. 2020)
- ingestion
- chunking strategies (size, overlap, structure-aware)
- metadata
- vector databases/indexes (pgvector, Cloudflare Vectorize, others — awareness)
- approximate nearest neighbour (HNSW idea)
Words you will know
What you will practise
- index the site's own lessons and query them
Interview questions this answers
- How do you choose a chunk size?
- How do you evaluate a RAG system?