Part IX — Applications and hybrid retrieval
These are the topics in this section, each on its own page with a stable path you can bookmark or share.
What topics are covered in this section?
- How does filtered and hybrid search change what HNSW has to do?
- How is HNSW used inside recommendation systems?
- How is HNSW used for image and multimodal search?
- How is HNSW used for semantic search and retrieval-augmented generation?
- What extra machinery does HNSW need once it’s inside a vector database?
Part IX connects HNSW to products: recommendations, images and multimodal search, semantic search and RAG, vector-database machinery, and filtered or hybrid retrieval.
What changes when HNSW leaves the lab?
Applications impose filters, freshness, multi-tenancy, hybrid lexical signals, chunking strategies, and rerankers. The graph becomes one stage in a pipeline. This Part shows those couplings so you tune for user outcomes, not only recall@10 on SIFT.
Five chapters move from recommenders, to vision, to RAG, to databases, to filtered/hybrid search.
How does Weaviate appear here?
Weaviate is the reference vector database illustrating CRUD HNSW, hybrid search, filters, and tenancy – always within the site rule that only Weaviate is named among products. Applications and filtered-search glossaries mirror the chapters.
Read your domain chapter closely, then the database and filter chapters every app eventually needs.
Where should you go next?
Pick the chapter for your use case first, then read vector-database and filtered/hybrid chapters. Continue to Part X for updates and failure modes in living systems.
Part IX is HNSW in production narratives. Next, open the chapter that matches your product, then the database machinery chapter before you harden operations.