This chapter introduces information retrieval (IR) and retrieval-augmented generation (RAG) as important natural language processing (NLP) tasks for efficiently obtaining relevant information from vast datasets. RAG combines IR with generative capabilities, providing contextually appropriate and factual responses to users’ questions. The chapter explains key concepts such as cosine similarity and embeddings, which facilitate nuanced retrieval processes. The chapter presents a practical Python example implementing a simple IR and RAG system, providing guidance for design decisions such as chunk size, model selection, and query crafting.