Ragtime — A modular RAG pipeline (alpha)

 ██████╗  █████╗  ██████╗████████╗██╗███╗   ███╗███████╗
 ██╔══██╗██╔══██╗██╔════╝╚══██╔══╝██║████╗ ████║██╔════╝
 ██████╔╝███████║██║  ███╗  ██║   ██║██╔████╔██║█████╗
 ██╔══██╗██╔══██║██║   ██║  ██║   ██║██║╚██╔╝██║██╔══╝
 ██║  ██║██║  ██║╚██████╔╝  ██║   ██║██║ ╚═╝ ██║███████╗
 ╚═╝  ╚═╝╚═╝  ╚═╝ ╚═════╝   ╚═╝   ╚═╝╚═╝     ╚═╝╚══════╝

A modular RAG pipeline for the French government, powered by the Albert API.

Prerequisites

• An Albert API key — request one here
• curl (pre-installed on macOS / Linux / WSL / Git Bash)

Install

Linux / macOS / WSL / Windows (Git Bash):

curl -fsSL https://raw.githubusercontent.com/etalab-ia/ragtime/main/install.sh | bash

Quick Start

# Create your RAG project (prompts for API key and preferences)
ragtime setup mon-projet

# Start your app
cd mon-projet && just run

Your app opens at http://localhost:8000 — upload documents and ask questions.

# Chat with the interactive RAG learning assistant
just learn

Upgrade

Re-run the installer to get the latest version:

curl -fsSL https://raw.githubusercontent.com/etalab-ia/ragtime/main/install.sh | bash

Documentation

Guide	Description
Getting Started	Detailed installation, available commands, advanced setup
Understanding the RAG Pipeline	What each stage of the pipeline does and why it matters
ragtime.toml Reference	Every configuration option, presets comparison, environment overrides
Evaluation Guide	Generate synthetic datasets and measure RAG quality
Components Reference	Albert Client SDK, frontend apps, and modules
Windows Setup	Installation on Windows via Git Bash
Proxy & Network Setup	Install behind corporate proxies and VPNs

Contributing

Want to contribute to Ragtime itself? See CONTRIBUTING.md for the architecture overview and development setup.

License

See LICENSE for details.