RaGenie

An AI augmentation platform combining multiple LLM providers with advanced Retrieval Augmented Generation

Open Source · MIT License

What is RaGenie?

RaGenie is an open source AI augmentation platform that combines Retrieval Augmented Generation with multiple LLM providers. It helps professionals work smarter by providing context-aware AI assistance for research, writing, coding, and complex problem-solving.

The name cleverly combines "RAG" (Retrieval Augmented Generation) with "Genie"—your AI genie ready to assist with knowledge work. RaGenie evolved from Ragbot, representing a significant architectural evolution from a simple chatbot to a sophisticated agentic AI platform.

RaGenie supports OpenAI (GPT-4o, o1, o3), Anthropic (Claude Sonnet 4.5, Claude Opus 4.5), and Google (Gemini 2.5) through a unified LLM gateway, with automatic file monitoring, vector embeddings, and semantic search across your knowledge base.

Key Features

Multi-Provider LLM Gateway

Unified interface to OpenAI, Anthropic, and Google models. Switch providers seamlessly without changing your workflow.

Advanced RAG with Vector Search

Automatic file monitoring and indexing with Qdrant vector database. Semantic search finds relevant context across your entire knowledge base.

Agentic AI Workflows

LangGraph integration enables multi-step reasoning with a retrieve-augment-generate pipeline. Stateful conversation management with real-time streaming.

Microservices Architecture

Seven specialized services for auth, users, documents, conversations, LLM gateway, file watching, and embeddings. Horizontally scalable and production-ready.

Enterprise-Grade Infrastructure

PostgreSQL for persistence, Redis for caching, MinIO for object storage. Prometheus metrics and Grafana dashboards for monitoring.

Privacy-First, Self-Hosted

Your data stays yours. Deploy on your own infrastructure with Docker or Kubernetes. No vendor lock-in, full data sovereignty.

Architecture

RaGenie is built as a modern microservices platform:

Service Purpose
Auth Service User registration, login, JWT tokens
User Service Profile management, preferences
Document Service File handling, embeddings, ragbot-data API
Conversation Service Chat management, RAG context, LangGraph workflows
LLM Gateway Unified interface to all LLM providers
File Watcher Automatic file monitoring and indexing
Embedding Worker Asynchronous embedding generation

The backend is built with FastAPI (Python), with a React/TypeScript frontend in development. Infrastructure includes PostgreSQL, Redis, Qdrant, MinIO, and Nginx, all orchestrated with Docker Compose.

The Ragbot Ecosystem

RaGenie is part of a family of tools for AI-augmented knowledge work:

Use Case Recommendation
Quick setup, CLI-focused workflow Ragbot
Prefer Streamlit simplicity Ragbot
Need advanced RAG with vector search RaGenie
Need microservices architecture RaGenie
Want both CLI and modern web UI Use both!

Ragbot is the original AI assistant with a CLI and Streamlit interface—perfect for quick setup and command-line workflows. Both products share the same data layer and continue to be actively developed.

Resources