Parent-Child Chunking: Advanced Document Splitting for RAG

Parent-child chunking tutorial for RAG: small-to-big retrieval with LlamaIndex HierarchicalNodeParser, AutoMergingRetriever, and LangChain ParentDocumentRetriever.

Contextual Compression in RAG: Retrieve Less, Answer Better

Contextual compression RAG tutorial with LangChain: LLMChainExtractor, EmbeddingsFilter, reranker pipelines, and measuring gains with RAGAS context precision.

RAGAS Tutorial: Automated Evaluation for RAG Pipelines

RAGAS tutorial: evaluate RAG pipelines with faithfulness, answer relevancy, context precision, and context recall metrics using LLM-as-a-judge and synthetic testsets.

Haystack Tutorial: Build a RAG Pipeline with Deepset's Framework

Haystack 2.x tutorial: build a full RAG pipeline with indexing (DocumentSplitter, embedder, writer) and query (retriever, PromptBuilder, OpenAIGenerator) components from deepset.

Supabase Vector Tutorial: pgvector Made Easy

Supabase pgvector tutorial: enable the vector extension, create a VECTOR table, add an HNSW index, generate embeddings, and build semantic search with match_documents in SQL.

Redis as a Vector Database: Fast In-Memory Semantic Search

Redis vector search tutorial with RedisVL: HNSW indexing, FT.HYBRID hybrid search, tag filtering, and semantic caching for LLM cost reduction.

Elasticsearch for Vector Search: kNN and Dense Retrieval

Elasticsearch vector search tutorial: dense_vector kNN, ELSER sparse retrieval, pre-filtering, and single-request BM25 + sparse + dense RRF fusion.

Milvus Tutorial: Scalable Vector Database for Large Datasets

Hands-on Milvus tutorial: Milvus Lite, collections, embedding model, filtering, Standalone migration, and GPU-accelerated billion-scale vector search.

Qdrant Tutorial: Open-Source Vector Search Engine Getting Started

Hands-on Qdrant tutorial: install the client, run in-memory or Docker, create collections, FastEmbed, payload filtering, hybrid search, and INT8 quantization.

HyDE Explained: Hypothetical Document Embeddings for RAG

HyDE explained: Hypothetical Document Embeddings close the query-document gap with LangChain HypotheticalDocumentEmbedder, custom prompts, LlamaIndex transforms, and RAGAS measurement.