Skip to content

jayeshgit65/QueryAI-AI-Powered-Multi-Source-Knowledge-Assistant

Folders and files

NameName
Last commit message
Last commit date

Latest commit

Β 

History

4 Commits
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 

Repository files navigation

QueryAI - AI Powered Multi-Source Knowledge Assistant

QueryAI is a cutting-edge AI application that allows users to interactively query and gain insights from multiple data sources, including PDFs and websites. By leveraging advanced embeddings and retrieval-augmented generation (RAG), QueryAI provides accurate, context-aware responses to user queries in real-time.


πŸš€ Features

  • Multi-Source Knowledge Integration: Combine PDFs and web content into a unified knowledge base for seamless querying.
  • Real-Time Conversational AI: Ask questions and receive contextually relevant answers using Google Gemini LLM.
  • Vector Database Powered: Efficient storage and retrieval of information using FAISS vector store with HuggingFace embeddings.
  • Chunking & Retrieval: Automatically splits large documents into manageable chunks for better retrieval and context management.

πŸ› οΈ Technology Stack

  • Backend & AI: Python, LangChain, FAISS, HuggingFace Embeddings, GoogleGenerativeAI
  • Document Processing: PyPDF2 for PDFs, WebBaseLoader for URLs
  • Web App: Streamlit for interactive user interface
  • Environment Management: dotenv for environment variables

πŸ’‘ How It Works

  1. Source Input: Users provide one or more PDFs or URLs via the Streamlit sidebar.
  2. Processing & Chunking: PDFs and web content are processed into text chunks for embeddings.
  3. Vector Store Creation: Chunks are converted into embeddings and stored in a FAISS vector store for efficient retrieval.
  4. Query & Response: Users ask questions in natural language. The RAG pipeline retrieves relevant context and generates an AI response.

πŸ—οΈ Architecture Diagram

QueryAI Architecture


⚑ Key Highlights

  1. Multi-Source Knowledge Retrieval: Process and query multiple documents and websites simultaneously for enriched responses.
  2. Real-Time Interaction: Instantaneous conversational experience with context-aware responses, enabling rapid decision-making.
  3. Persistent Vector Store: FAISS vector store allows scalable and efficient retrieval across large datasets.

πŸ“‚ Usage

  1. Clone the repository:
git clone https://github.com/jayeshgit65/QueryAI-AI-Powered-Multi-Source-Knowledge-Assistant
cd QueryAI-AI-Powered-Multi-Source-Knowledge-Assistant
  1. Install dependencies:
pip install -r requirements.txt
  1. Create a .env file and configure your API keys (if any).

  2. Run the Streamlit app:

streamlit run app.py
  1. Add PDF or URL sources in the sidebar, process them, and start querying!

πŸ“ˆ Impact

  • Enables data-driven insights from heterogeneous sources in one platform.
  • Improves productivity by providing accurate, contextual answers in real-time.
  • Suitable for researchers, analysts, and enterprises to query internal and external knowledge sources efficiently.

πŸ”— License

This project is licensed under the MIT License.

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages