Skip to content
Discussion options

You must be logged in to vote

RAG stands for Retrieval-Augmented Generation.

In simple terms, RAG is a way to make AI answers smarter and more accurate by letting the model look up information before generating a response.

Why RAG exists

Large Language Models (LLMs) like ChatGPT:

  • Don’t know private data

  • May have outdated knowledge

  • Can hallucinate (make things up)

RAG solves this by combining search + generation.

How RAG works (step by step)

  1. User asks a question
  • “What are the features of our Book Shop Management System?”
  1. Retriever searches data
  • Database

  • PDFs

  • Docs

  • Knowledge base

  • Vector database (embeddings)

  1. Relevant information is fetched
  • Only the most related content is selected
  1. Generator…

Replies: 1 comment 1 reply

Comment options

You must be logged in to vote
1 reply
@kelumwanninayaka
Comment options

Answer selected by kelumwanninayaka
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Category
Q&A
Labels
None yet
2 participants