RAG

This repo will showcase how to use vector RAG and use other techniques to overcome its weaknesses.

Vector RAG

Store private data in vector DB

• Get text of some domain
• store in vector DB

Chat Interface

• Fastest to use openAI API with a small model

Retrival

• ANN or other similarity measure with vector DB

Augment

• Append context to prompt

Generate reponse

• Produce output