Vector Database

Detail a kontext

Vector Database (Vector DB) ukládá data jako vysoce-dimensional vektory (embeddings) místo klasických rows. Umožňuje similarity search — místo exact match WHERE name = "X" hledáte "podobné". Použití v marketingu: RAG systémy (čat AI s vašimi dokumenty), semantic search na webu, product recommendations, content similarity, chatbot knowledge bases. Populární: Pinecone (managed, dominant), Qdrant (open-source self-hosted), Weaviate (open-source + cloud), Chroma (lightweight, Python-native), pgvector (PostgreSQL extension). Embeddings: OpenAI text-embedding-3-small ($0.02/M tokens), Cohere, open-source sentence-transformers. Setup pro RAG: chunk documents → embed → store → query → retrieve top-K → feed LLM.