Document Chat

Transform documents into intelligent conversations with AI. Open-source document chat platform with semantic search, RAG (Retrieval Augmented Generation), and multi-provider AI support (OpenRouter, OpenAI, ImageRouter).

🌐 Test it out: https://document-chat-system.vercel.app

💻 GitHub Repo: https://github.com/watat83/document-chat-system

🎥 Watch Video Explainer: https://youtu.be/P42nlCmicVM?si=maIjXVxaKWkvevn9

Overview Document Chat System is a full-stack, production-ready application that combines intelligent document management with AI-powered conversations. Upload documents in multiple formats, automatically process and index them, then have natural language conversations about your content using advanced AI models.

Key Features

🆓 100% Free & Open Source - MIT licensed. Deploy your own instance, modify as needed, or monetize as a SaaS.

🤖 Multi-Provider AI - Supports OpenRouter (100+ models), OpenAI, Anthropic, and ImageRouter. Uses gpt-4o-mini by default for cost-effective responses.

📄 Full Document Support - PDFs, DOCX, TXT, images with OCR, and more. Automatic text extraction and intelligent processing.

🔍 Semantic Search - Vector search with Pinecone or pgvector finds relevant content beyond simple keyword matching.

👥 Multi-Tenant Ready - Built-in organization isolation with complete data separation between users/organizations.

💳 Optional SaaS Billing - Integrated Stripe billing system with customizable pricing plans for monetization.

⚡ Background Processing - Inngest handles document processing, vectorization, and AI analysis asynchronously.

🎨 Modern UI - Beautiful, responsive interface with dark mode built using shadcn/ui and Tailwind CSS.

🔐 Enterprise Security - AES-256 encryption, Row-Level Security (RLS), and Clerk authentication.

🐳 Production Ready - Dockerfile included, deploy to Vercel/Railway/Render in minutes with one-click setup.

Features

📁 Document Management Multi-Format Support: PDF, DOCX, TXT, MD, images, and more formats will be supported soon Intelligent Processing: Automatic text extraction, OCR, metadata analysis Folder Organization: Hierarchical folder structure with drag-and-drop Batch Operations: Upload and process multiple files simultaneously Real-Time Progress: Live updates on document processing status

🤖 AI-Powered Chat Multiple AI Providers:

• OpenRouter: Access to 100+ models (GPT-4, Claude, Llama, Mistral, etc.)
• OpenAI: Direct integration with GPT-4 Turbo and GPT-3.5
• ImageRouter: Visual AI for image analysis and OCR
• Document Context: AI understands and references your uploaded documents
• Source Citations: Responses include references to source documents
• Streaming Responses: Real-time token streaming for faster interactions

🔍 Advanced Search

• Vector Search: Semantic search powered by Pinecone or PostgreSQL pgvector
• Hybrid Search: Combines semantic similarity with keyword matching
• Full-Text Search: Fast text search across all documents
• Filters: Filter by date, type, folder, tags, and more

👥 Multi-Tenant Architecture

• Organization Isolation: Complete data separation between organizations
• Per-Org Resource Limits: Customizable limits per organization
• Activity Tracking: Audit logs for compliance and security

⚡ Background Processing

• Inngest Integration: Event-driven serverless functions
• Document Processing Queue: Scalable batch processing
• Automatic Retries: Built-in error handling and retries
• Real-Time Notifications: Progress updates via webhooks
• Job Monitoring: Track job status and logs

License This project is licensed under the MIT License - see the LICENSE file for details.

What this means:

✅ Commercial use allowed

✅ Modification allowed

✅ Distribution allowed

✅ Private use allowed

❌ No liability

❌ No warranty