title	DocQuery - RAG Document Assistant
emoji	📄
colorFrom	blue
colorTo	green
sdk	gradio
sdk_version	5.50.0
app_file	app.py
pinned	true

📄 DocQuery - RAG Document Assistant (Groq + LCEL + Pinecone)

Upload one or more PDF documents and ask questions. Fully cloud-based with Pinecone vector DB 🧠 and powered by Llama 3 via Groq API with LCEL-style prompting.

Features

• Retrieve answers from multiple PDF documents using RAG (Retrieval-Augmented Generation)
• Top-K chunk retrieval ensures relevant context is used
• Fast, actionable, and professional AI responses
• Embeddings powered by sentence-transformers/all-MiniLM-L6-v2
• Supports multiple PDF document uploads and queries
• Persistent memory across sessions
• chat interface with retry functionality

Usage

1. Upload PDF files.
2. Enter your question in the textbox.
3. Click submit to get answers.
4. Click retry to get answers for the last question.

Notes

• The .env file containing your Groq API Key and Pinecone API Key must not be committed to the repository.
• Ensure all dependencies in requirements.txt are installed.

Dependencies

• gradio
• langchain-community
• pinecone-client
• langchain-groq
• sentence-transformers
• python-dotenv

---

Powered by onwurahben