Category:

Retrieval-Augmented Generation (RAG)

Category:

RAG & Retrieval

Definition

Combining LLMs with external knowledge retrieval to improve accuracy.

Explanation

Retrieval-Augmented Generation enhances LLM outputs by retrieving relevant documents or data before generation. This grounds responses in enterprise knowledge, reduces hallucinations, and enables up-to-date answers without retraining models.

Technical Architecture

Query → Retrieval → Context Injection → LLM Generation → Output

Core Component

Retriever, embeddings, vector database, LLM

Use Cases

Enterprise search, knowledge assistants, analytics QA

Pitfalls

Poor retrieval quality, incorrect chunking

LLM Keywords

RAG, Retrieval Augmented Generation

Related Concepts

Related Frameworks

• Embeddings
• Vector Database
• Chunking

• LangChain RAG
• LlamaIndex

Back to Glossary Index