README.md

metadata

title: Multimodal RAG Service
emoji: 🚀
colorFrom: blue
colorTo: purple
sdk: docker
app_port: 7860
pinned: false

🚀 Multimodal RAG Service

A unified service for PDF ingestion and multimodal querying using RAG (Retrieval-Augmented Generation).

Features

• 📄 PDF text extraction
• 📊 Table extraction
• 🖼️ Image extraction
• 🤖 Multimodal summarization
• 🔍 Vector similarity search
• 💬 Context-aware answering

API Endpoints

• POST /ingest - Upload and process PDF files
• POST /query?question=YOUR_QUESTION&k=5 - Query the system
• GET /stats - View system statistics
• GET /docs - Interactive API documentation

Usage

Visit the deployed space and use the /docs endpoint for interactive API documentation.