myRAG - Multi-user RAG System

A high-performance, self-hosted RAG (Retrieval-Augmented Generation) system with knowledge graph support.

Features

• 📚 Knowledge Bases - Create isolated knowledge bases per user/tenant
• 📄 Multi-format Support - PDF, DOCX, XLSX, Markdown, CSV, TXT
• 💬 AI Chat - Chat with your documents, get answers with citations
• 🔍 Hybrid Search - Vector + keyword search with optional GraphRAG
• 👥 Multi-tenant - Row-level security for data isolation
• 🚀 High Performance - Go backend with async document processing
• 🐳 One-click Deploy - Docker Compose setup

Quick Start

# Clone the repository
git clone https://github.com/njujiangxiang/myRAG.git
cd myRAG

# Start all services
docker-compose up -d

# Access the web UI
open http://localhost:3000

Architecture

┌─────────────┐     ┌─────────────┐     ┌─────────────┐
│   React     │────▶│   Go API    │────▶│  PostgreSQL │
│   Frontend  │     │   Backend   │     │  (Metadata) │
└─────────────┘     └──────┬──────┘     └─────────────┘
                           │
                    ┌──────▼──────┐
                    │   NATS JS   │
                    │  (Queue)    │
                    └──────┬──────┘
                           │
                    ┌──────▼──────┐
                    │   Workers   │
                    └──────┬──────┘
                           │
         ┌─────────────────┼─────────────────┐
         ▼                 ▼                 ▼
┌─────────────┐    ┌─────────────┐   ┌─────────────┐
│   Qdrant    │    │    MinIO    │   │  LLM APIs   │
│  (Vectors)  │    │   (Files)   │   │ (OpenAI)    │
└─────────────┘    └─────────────┘   └─────────────┘

Tech Stack

• Backend: Go 1.21+ with Gin framework
• Frontend: React + TypeScript + shadcn/ui + Tailwind CSS
• Database: PostgreSQL 16
• Vector DB: Qdrant 1.7+
• Message Queue: NATS JetStream
• File Storage: MinIO (S3-compatible)
• LLM: OpenAI/Anthropic/Local (ollama)

Configuration

Copy .env.example to .env and configure:

# Database
DATABASE_URL=postgres://user:pass@localhost:5432/ragdb

# Qdrant
QDRANT_URL=http://localhost:6333

# NATS
NATS_URL=nats://localhost:4222

# MinIO
MINIO_ENDPOINT=localhost:9000
MINIO_ACCESS_KEY=minioadmin
MINIO_SECRET_KEY=minioadmin

# LLM
OPENAI_API_KEY=sk-...

Development

# Run backend
go run cmd/server/main.go

# Run frontend (in web/ directory)
npm run dev

# Run tests
go test ./...

Roadmap

• v1.0: Core RAG functionality (vector search + chat)
• v1.1: GraphRAG integration (knowledge graph + entity extraction)
• v1.2: Advanced search (hybrid search + BM25 + reranking)
• v1.3: Frontend UI (multi-tenant dashboard)
• v1.4: SSE streaming for chat responses
• v2.0: Advanced permissions and team collaboration

License

MIT