RAGify Search - Intelligent Assistant with Web Search RAG

Description

RAGify Search is an AI-powered intelligent assistant built with Streamlit, designed to answer user queries by integrating real-time web search. It uses a Retrieval-Augmented Generation (RAG) approach to provide accurate, concise, and context-aware responses

---

Key Features

• Real-Time Web Search: Fetch relevant web pages for user queries.
• Document Processing: Split, embed, and index documents for similarity-based searches.
• Prompt Engineering: Generate optimized prompts for the LLM model.
• Streamlit Interface: Simple and intuitive chatbot-like user experience.
• Temporary File Management: Automatic cleanup of downloaded content to save space.
• Local LLM Support: Utilizes Ollama to run a local LLM, ensuring data privacy and high performance.

---

Directory Structure

• app.py: Front-end application powered by Streamlit.
• config.py: Centralized configuration for chunk sizes, model names, and directories.
• db_operations.py: Embeds and indexes document chunks for search.
• extract_queries.py: Generates optimized queries to be searched.
• web_scraper.py: Manages web scraping and temporary file handling.
• prompt_generator.py: Prepares structured prompts for context-aware answers.

---

Installation

1. Clone the Repository:

   git clone https://github.com/pcastiglione99/RAGify-Search.git
   cd RAGify-Search

2. Install Dependencies:

   pip install -r requirements.txt

3. Run the Application:

   streamlit run ./src/app.py

4. langchain ollama requirement

   ollama pull nomic-embed-text

---

Usage

1. Start the Streamlit app and interact with the chatbot.
2. Enter your query in the chat interface.
3. The assistant fetches web pages, processes content, and generates a response.