🚀 RAG Me Up - by SensAI.PT

A simple and extensible framework to build RAG (Retrieval-Augmented Generation) applications fast.

⚡ TL;DR – Installation & Quickstart

# Clone the repo
git clone https://github.com/SensAI-PT/RAGMeUp.git
cd RAGMeUp

# Create and populate your Docker env file
cp docker-compose.env.example docker-compose.env
# Edit docker-compose.env and set at least POSTGRES_PASSWORD and JWT_SECRET

# Build and start everything
docker compose --env-file docker-compose.env up --build -d

React UI is available on http://localhost (or the HOST_PORT you set in docker-compose.env).

Note: The full Docker Compose setup runs the Python server in CPU-only mode (no GPU/CUDA access inside Docker). If you need GPU acceleration for embeddings and inference, use the hybrid mode below.

🖥️ Hybrid Mode – GPU Support

In hybrid mode, Postgres, the Node.js API server, and the React client run in Docker, while the Python RAG server runs standalone on the host where it has full access to your GPU/CUDA.

1. Start the Docker services (without the Python server)

cp docker-compose.env.example docker-compose.env
# Edit docker-compose.env — uncomment and adjust PYTHON_SERVER_URL and POSTGRES_PORT if needed

docker compose --env-file docker-compose.env \
  -f docker-compose.yml -f docker-compose.hybrid.yml up --build -d

This starts: - ParadeDB Postgres – exposed on the host at localhost:6024 (configurable via POSTGRES_PORT) - Node.js API server – connects to your host Python server via PYTHON_SERVER_URL (default: http://host.docker.internal:5000) - React client (nginx) – accessible at http://localhost:HOST_PORT

2. Run the Python RAG server on the host

cd server

# Set up a virtual environment (first time only)
python -m venv .venv
# Linux/macOS:
source .venv/bin/activate
# Windows:
.venv\Scripts\activate

# Install dependencies (first time only)
pip install -r requirements.txt

# Make sure server/.env has:
#   postgres_uri="postgresql://langchain:langchain@localhost:6024/langchain"
#   embedding_cpu=False   (to use GPU)

python server.py

The Python server starts on port 5000 by default. The Dockerized Node server reaches it via host.docker.internal:5000.

Windows/macOS: host.docker.internal works out of the box. Linux: Add --add-host=host.docker.internal:host-gateway to each service in the compose file, or set PYTHON_SERVER_URL=http://172.17.0.1:5000 in docker-compose.env.

📘 Documentation

Full setup instructions, architecture docs, API references, and guides available at:

👉 https://ragmeup.sensai.pt

🧠 Why RAG Me Up?

⚙️ Modular: Use your own chunkers, vectorstores, or retrievers

🚀 Fast to prototype: Focus on your RAG logic, not boilerplate

🧩 Flexible: Plug-and-play architecture

✨ Battle-tested: RAG Me Up has been used in many large-scale production settings, most notably in SensAI.PT - Your AI personal trainer

🤝 Contributing

We welcome pull requests, feedback, and ideas. Open an issue or start a discussion to get involved.