RAGify – Chat with Your Documents, Anytime

RAGify is an intelligent retrieval-augmented chat application that lets users upload documents, ask questions about their content, and manage a personal knowledge base with persistent conversation history.

Overview

This project implements the RAG (Retrieval-Augmented Generation) pattern to create a chatbot that answers questions based on a set of documents provided by the user. The application is fully interactive, with a web interface built in Streamlit, and includes a robust backend to manage users, conversations, and each user's vector indexes.

---

🔄 Application Flow

The diagram below illustrates the flow of the application, from user login to document upload, processing, and interaction with the chat.

Application Flow

---

✨ Key Features

• User Authentication: Account Sign Up/Login system to ensure that each user’s data is private and persistent.
• Multi-format Upload: Supports uploading .pdf, .docx, .xlsx, .csv, .txt, and .md files.
• User Data Persistence:
  • Chat History: Conversations are saved in a SQLite database and reloaded on each login.
  • Knowledge Base: Uploaded documents are converted into vectors and stored in a dedicated FAISS index for each user.
• File Management: Users can view and delete previously uploaded files, with the knowledge base updated accordingly.

---

🛠️ Technologies Used

• Streamlit
• Python 3.12+
• LLM and RAG: Ollama with the llama3 model and LangChain.
• Relational Database: SQLite
• Vector Database: Facebook AI Similarity Search (FAISS)
• Embeddings Model: nomic-ai/nomic-embed-text-v1 from Hugging Face

---

🚀 Installation and Setup

Follow the steps below to set up and run the project locally.

1. Prerequisites

• Python 3.12 or higher: Download Python
• Git: To clone the repository.
• Ollama: The application uses Ollama to run the Llama 3 model locally.
  • Install Ollama.
  • After installation, pull the llama3 model with:

    ollama pull llama3

  • Make sure Ollama is running before starting the Streamlit app.

2. Clone the Repository

git clone https://github.com/PLeonLopes/RAGify.git
cd RAGify

3. Create a Virtual Environment and Install Dependencies

# Create the virtual environment
python3 -m venv venv

# Activate the virtual environment
venv\Scripts\activate           # <- Windows

source venv/bin/activate        # <- macOS/Linux

Now install all dependencies with:

pip install -r requirements.txt

4. Run the Application

With the virtual environment activated, run Ollama and start the Streamlit app:

Start llama3

ollama serve

Start Streamlit Application

streamlit run src/app.py

The application should automatically open in your default browser.

📖 How to Use

1. Create an Account: In the sidebar, enter a username and password and click "Create Account".

2. Login: Use the same credentials to log in.

3. Upload Files: In the sidebar, select one or more documents and click the "Process Files" button.

4. Wait for Processing: The application will extract the text, generate embeddings, and save your knowledge base.

5. Chat: Once the files are processed, go to the main chat area, type your question, and click "Send".

6. Manage Your Files: In the sidebar, you can view the list of uploaded files and remove any of them by clicking the trash icon.

Authors

Pedro Lopes

Vinícius Vieira