Retrieval-augmented generation (RAG) has emerged as a transformative approach to enhancing the capabilities of large language models (LLMs) by integrating real-time information retrieval with generative text synthesis. Traditional LLMs rely on pre-trained data, limiting their effectiveness when handling tasks requiring up-to-date or domain-specific information. RAG addresses this constraint by embedding a retrieval mechanism within the LLM framework, allowing the model to access external knowledge sources dynamically. This integration enables LLMs to generate contextually relevant and factually accurate responses, extending their utility within the realm of information systems (IS), where timely and precise information is critical. Starting from the history and evolution of retrieval-augmented generation, passing through different framework and methodologies, and moving to practical applications by different libraries, in this chapter, we offer a theoretical and practical overview of this innovative and powerful technique.