Artificial Intelligence
Retrieval-Augmented Generation (RAG)
Also known as: RAG & Retrieval-Augmented Generation
Plain English
AI that looks things up before answering.
Definition
RAG is a technique that combines glossaryInformation RetrievalInformation retrieval is the process of finding relevant information from large datasets based on a query or input.Open glossary term with glossaryGenerative AIGenerative AI refers to systems that create new content such as text, images, or code based on learned patterns from data.Open glossary term, allowing glossaryModelA model is a system or representation used to process data and generate outputs, often trained to perform specific tasks.Open glossary term to use external data to produce more accurate and context-aware responses.
In practice
Used to improve glossaryAI OutputAI output refers to any result generated by an AI system, including text, images, predictions, or decisions.Open glossary term by grounding glossaryResponseA response is the data or result returned by a server after receiving a request.Open glossary term in real glossaryDataData is raw, uninterpreted information collected and stored so it can be analysed, processed, or used to inform decisions. On its own it carries no meaning; context and interpretation are what make it useful.Open glossary term rather than relying solely on pre-trained knowledge.
In context
It is what makes an assistant able to answer from your documentation rather than its training. Most of the engineering effort turns out to be retrieval quality, not generation.
The reality
Without retrieval, AI guesses. With it, glossaryResponseA response is the data or result returned by a server after receiving a request.Open glossary term become more reliable and useful.
Compared with
RAG vs Fine-tuning
RAG fetches relevant material and gives it to the glossaryModelA model is a system or representation used to process data and generate outputs, often trained to perform specific tasks.Open glossary term at glossaryRequestA request is an action sent from a client to a server asking for data or a service.Open glossary term time, so answers can cite sources and update the moment the source does. guideFine-tuningWhat fine-tuning does to an AI model, when it is worth doing, and what product and design teams need to know before commissioning it.Open guide embeds knowledge in the weights, where it is faster but frozen and unattributable.
FAQ
Common questions
A few practical answers to the questions that usually come up around this term.
What is RAG?
RAG combines retrieval and AI generation to improve glossaryResponseA response is the data or result returned by a server after receiving a request.Open glossary term.
Why is RAG important?
It makes glossaryAI OutputAI output refers to any result generated by an AI system, including text, images, predictions, or decisions.Open glossary term more accurate and grounded.
Where is RAG used?
In AI glossarySystemA system is a collection of interconnected components that work together to achieve a specific function or outcome.Open glossary term, glossarySearchSearch is the functionality that allows users to find content or information by entering queries. It relies on indexing, metadata, and relevance algorithms to return useful results.Open glossary term, and knowledge tools.
What problem does RAG solve?
It reduces hallucination by using real glossaryDataData is raw, uninterpreted information collected and stored so it can be analysed, processed, or used to inform decisions. On its own it carries no meaning; context and interpretation are what make it useful.Open glossary term.
Related Services
Related Guides
Related Terms
