Skip to content
Codeloom

Topics / Embeddings & RAG

Embeddings & RAG

Retrieval-augmented generation. Embeddings, vector databases, and pipelines.

Browse 27 tutorials → Other topics

Why learn Embeddings & RAG?

  • Lets LLMs answer questions about your data, not just the open web.

  • The dominant pattern for AI chat products in 2026.

  • Vector search has applications well beyond chat.

  • A practical, high-impact skill for backend and AI engineers.

What you can build with Embeddings & RAG

Customer support over a knowledge base Internal "ask my docs" tools Semantic search and recommendations Multi-step agentic retrieval

Embeddings & RAG tutorials

27 articles · page 1 of 2

Hand-written tutorials, ordered as a recommended learning path.

  1. 01 What Is RAG? A clear introduction to retrieval-augmented generation — why LLMs don't know your data, how embeddings and vector search solve it, and when RAG beats fine-tuning.
  2. 02 Embeddings What an embedding is, why cosine similarity works, how dimensionality and chunking choices affect retrieval, and a tiny numpy example you can run in your head.
  3. 03 Vector Databases Why a normal database struggles with vector search, how ANN indexes like HNSW and IVF work, and a clear comparison of pgvector, Qdrant, Pinecone, Chroma, and Weaviate so you can pick one.
  4. 04 Chunking Strategies Learn fixed-size, semantic, recursive, and document-aware chunking strategies for RAG pipelines. Includes overlap techniques, benchmarks, and Python code.
  5. 05 Embedding Models Compare text-embedding-3-small/large, Cohere embed-v3, and sentence-transformers on dimensions, cost, MTEB scores, and practical RAG performance.
  6. 06 Hybrid Search Learn how to combine BM25 keyword search with vector similarity using reciprocal rank fusion for better RAG retrieval. Includes Python code and benchmarks.
  7. 07 Evaluation Metrics Learn to evaluate RAG pipelines with Recall@k, MRR, NDCG for retrieval and faithfulness, relevance, hallucination rate for generation. Includes RAGAS setup.
  8. 08 Advanced Retrieval Improve RAG retrieval quality with advanced techniques: query expansion, multi-query retrieval, cross-encoder reranking, and reciprocal rank fusion.
  9. 09 RAG in Production Build a production-ready RAG pipeline with document ingestion, chunking, retrieval, and generation. Covers error handling, monitoring, and optimization.
  10. 10 Vector DB Comparison Compare Pinecone, Weaviate, and Chroma vector databases for RAG applications. Covers setup, performance, pricing, and when to use each one.
  11. 11 Context Windows Why bigger context windows are not always better: cost, attention degradation, retrieval design, and how to architect for long-context tasks.
  12. 12 Chunk Overlap Learn how chunk overlap rescues boundary context in RAG pipelines, with practical strategies for choosing overlap size and shape for different corpora.
  13. 13 Chunking Compare fixed-size, sentence, semantic, and structural chunking for retrieval augmented generation and pick the right one for your corpus.
  14. 14 Doc Loaders An overview of document loaders in RAG pipelines, covering common formats, libraries, and how to choose the right loader for your data.
  15. 15 RAG Eval Measure RAG quality with recall@k, MRR, context precision, faithfulness, and answer relevancy so you can iterate on data, not vibes.
  16. 16 Graph RAG Graph RAG combines knowledge graphs with retrieval augmented generation to handle multi-hop questions and complex reasoning.
  17. 17 Hybrid Search Combine lexical BM25 with dense vector search to recover the queries each method misses on its own and boost RAG retrieval quality.
  18. 18 HyDE Learn how Hypothetical Document Embeddings (HyDE) improve RAG recall by embedding a generated answer instead of the raw query, with examples and trade-offs.
  19. 19 Metadata Filters How to use metadata filters in RAG to improve precision, scope retrieval, and enforce permissions without sacrificing recall.
  20. 20 Rerankers Add a reranker on top of vector retrieval to dramatically improve top-k quality with cross-encoders, late interaction, and LLM rerankers.