This repository contains a collection of small projects designed to explore the application of RAG (Retrieval-Augmented Generation) in real-world scenarios.

The primary objective is to acquire hands-on experience and develop a deeper understanding of how to effectively utilize RAG in tackling practical problems

pip install -r requirements.txt

A simple question and answer application using LangChain framework and Chroma as the vector database.

The documents need to be in PDF format.

• Put your PDF files in the pdfs folder.
• Execute the script running the command python chatter.py.
• Make questions about the document's context.

Run the command python chatter.py to start the application.

Using QA LangChain module.

Works the same way as chatter.py, but using a different module.

Run the command python chatter-qa.py to start the application.

This is a example of how to use RAG to answer questions in a SQL database.

There are tow database in the db folder:

• chinook.db: A SQLite database about music store.
• movies.db: A SQLite database about movies store.

Run the command python chatter-sql.py to start the application.

A simples question and anwser application using LlamaIndex framework.

This application works the same way as chatter.py, but using a different framework keeping the chat history.

Run the command python chatter-llama-index.py to start the application.

• LangChain - Q&A with RAG
• LangChain - Q&A over SQL + CSV
• LlamaIndex - RAG Applications for Production