QueryLens

QueryLens is a cutting-edge document interaction platform that transforms static PDFs into interactive knowledge bases. By combining Retrieval-Augmented Generation (RAG) with advanced natural language processing it enables users to "converse" with their documents extracting insights with unprecedented precision.

🚀 Demo

https://www.loom.com/share/fcad98d57e0b4fbdafbf62ea7edd27bb?sid=dd222dd1-c7b2-40a4-83d5-5dc03e5ba896

Project Screenshots:

🧐 Features

Here're some of the project's best features:

• Ask Anything: Pose natural language questions about PDF content
• Context-Aware Answers: Leverages vector embeddings and Groq's ultra-fast LLMs for accurate citation-ready responses
• Drag-and-Drop PDF Processing: Automatic text extraction and vector embedding generation
• Semantic Search Engine: Pinecone-powered vector database enables content-based document retrieval
• In-App PDF Viewer: Annotate preview and download documents without leaving the platform
• Cross-Document Analysis: Ask questions spanning multiple uploaded PDFs

🛠️ Installation Steps:

1. Starting App

Double click "start-app file"

2. Enter Your API Keys

Copy and paste your API Keys

OR

2. Make a file named ".env" inside the server folder

GROQ_API_KEY=your_groq_key
PINECONE_API_KEY=your_pinecone_key
PINECONE_ENV=your_pinecone_region
HUGGINGFACEHUB_API_TOKEN=your_hf_token

💻 Built with

Technologies used in the project:

• React + Vite
• Lucid-React
• FastAPI
• Uvicorn
• Pydantic
• Groq
• LangChain
• Pinecone
• PyPDF
• Hugging Face
• Python-multipart
• FileResponse

🌊 Flow of the Tech Stack:

Frontend/Client → FastAPI Server → LangChain RAG Pipeline → Groq LLM
PDF Upload → PyPDF Loader → Text Splitters → Hugging Face Embeddings → Pinecone Vector Store
Query → Pinecone Retrieval → Groq LLM → Response