Vector index
A vector index is a data structure that organizes vector representations to support efficient similarity or nearest-neighbor search.
Machine Learning Platform Engineering
A vector index is an indexing structure that organizes numerical vector representations to support efficient similarity search.
- How it is used
- Enables fast semantic retrieval of document chunks for RAG prompts.
- How deep it goes
- Works through chunk preparation, embedding, FAISS creation, persistence, and querying; does not detail FAISS index internals.
- Examples
- Uses Markdown documents, a FAISS index, and an encryption-key question to test retrieved chunks.
- What you can do after
- Build it from scratch with FAISS, persist it, and reconnect it to a retriever for RAG queries.
Read how Machine Learning Platform Engineering explains vector index →
Generative AI in Action
A vector index is a data structure that organizes high-dimensional vectors to support efficient similarity and nearest-neighbor searches.
- How it is used
- Enables fast retrieval of similar document embeddings for RAG prompts rather than brute-force comparison across a large corpus.
- How deep it goes
- Defines the data structure and surveys indexing approaches, metrics, compression, and trade-offs without deriving their algorithms.
- Examples
- Uses a Redis-backed blog-post chat application that indexes chunks and retrieves five matches through hybrid search.
- What you can do after
- Operate/configure it by choosing indexing and similarity options for retrieval, then assessing speed, accuracy, memory, and production needs.
MongoDB 8.0 in Action, Third Edition
A vector index is a specialized index structure that organizes vector embeddings for efficient similarity search.
- How it is used
- Enables similarity retrieval over stored embeddings for semantic search, recommendations, and chatbot retrieval.
- How deep it goes
- Works through index configuration, CLI creation, query use, and performance choices without implementing an indexing algorithm.
- Examples
- Uses a movie-plot embedding index with filter fields and a chatbot embedding index verified as ready.
- What you can do after
- Operate/configure it with the Atlas CLI, then query the index through $vectorSearch.
Read how MongoDB 8.0 in Action, Third Edition explains vector index →
Build Python Web Apps with Streamlit
A vector index is a searchable Pinecone collection that stores text records, metadata, and their vector embeddings.
- How it is used
- Deploys a searchable knowledge base for retrieval-augmented customer-service applications.
- How deep it goes
- Explains index setup, document chunking, ingestion, retrieval, and prompt augmentation, but does not cover index maintenance or evaluation.
- Examples
- Uses pen-care support articles, a Pinecone dashboard, ingestion scripts, and a chatbot retrieving knowledge-base excerpts.
- What you can do after
- Operate/configure a Pinecone-backed vector store that ingests text files and retrieves relevant documents for an agent.
Read how Build Python Web Apps with Streamlit explains vector index →
Domain-Specific Small Language Models
A data structure that accelerates retrieval by organizing vectors for efficient similarity search.
- How it is used
- Accelerates similarity retrieval for embedding-based classification and retrieval-augmented generation.
- How deep it goes
- Explains index construction, normalization, distance results, and offline-online placement, but does not compare index algorithms.
- Examples
- Demonstrates Marvel sentence classification with FAISS and PDF-chunk retrieval with an embedding-backed RAG pipeline.
- What you can do after
- Apply it via FAISS by indexing normalized embeddings, searching query vectors, and mapping returned positions to source records.
Read how Domain-Specific Small Language Models explains vector index →
Natural Language Processing in Action, Second Edition
Vector search indexing is the process of organizing numerical vector representations so that semantically similar items can be retrieved efficiently.
- How it is used
- Supports question answering by retrieving relevant document vectors from a document store.
- How deep it goes
- Explains document conversion, store initialization, embedding updates, and Faiss index choices, but stops short of algorithm internals.
- Examples
- Uses AsciiDoc manuscript lines, Haystack Documents, FAISSDocumentStore, and Faiss Flat, HNSW, and IVF options.
- What you can do after
- Build it from scratch with Haystack by storing documents, updating embeddings, and saving a Faiss-backed index.
Read how Natural Language Processing in Action, Second Edition explains vector index →
AI-Powered Search
An approximate nearest-neighbor index is an index structure that efficiently retrieves vectors close to a query vector while allowing approximate results.
- How it is used
- Supports semantic retrieval and autocomplete by finding indexed content with vectors similar to a query vector.
- How deep it goes
- Shows index construction and ANN querying with NMSLIB and HNSW, then points production use toward search-engine APIs.
- Examples
- Uses outdoors concepts for autocomplete and hiking-question titles retrieved for the query “mountain hike.”
- What you can do after
- Build and query an ANN index over embeddings, or use a configured search engine's vector-search interface.
AI Agents in Action, Second Edition
An approximate nearest neighbor index accelerates similarity retrieval by finding vectors that are close to a query without exhaustively comparing every stored vector.
- How it is used
- Accelerates RAG retrieval as one option among complementary search and ranking techniques.
- How deep it goes
- Names ANN indexes as a retrieval optimization but does not explain their construction or search mechanics.
- Examples
- Shows a retrieval function configured for top-k passages with ANN search and metadata filtering.
- What you can do after
- Apply it via a retrieval tool to fetch top-k passages, with metadata filtering as a complementary option.
Read how AI Agents in Action, Second Edition explains vector index →
Related concepts
- similarity searchused by · Similarity search is a retrieval technique that finds or ranks items by comparing their representations with a query according to a similarity or distance measure.
- embeddinghas part · A numerical vector representation of an item or input that encodes features or relationships useful for computational tasks.
- Retrieverused by · A retriever is a software component that searches or queries an information source and returns items relevant to a query for use by downstream processes.
- vector databasepart of · A vector database is a data-management system that stores and indexes numerical vector representations and supports retrieval through similarity or distance searches.
- hybrid searchused by · Hybrid search is an information-retrieval technique that combines multiple retrieval methods or relevance signals to find or rank relevant results.
These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.