Qux360 Demo

This repository is a web demo for Qux360, an experimental Python library for AI-assisted qualitative analysis.

📦 Installation

.env files are needed for both the backend and the frontend. NOte that the underlying Qux360 library uses LiteLLM, i.e. you will need to specify your inference service in a way that is compatible with LiteLLM. Below is just an example for watsonx.

backend/.env:

MODEL_ID=watsonx/meta-llama/llama-3-3-70b-instruct
WATSONX_URL=[your URL]
WATSONX_API_KEY=[your API key]
WATSONX_PROJECT_ID=[your project ID]
DISABLE_AIOHTTP_TRANSPORT=True

frontend/.env:

VITE_BACKEND_URL=http://localhost:8000

👩🏻‍💻 Option 1: Run it in your local environment without Docker (dev)

Backend

cd backend

The first time you run it you need to run:

poetry install
source .venv/bin/activate

That will create a venv virtual environment for you.

In case you prefer to install it manually, the requirements.txt file is provided:

pip install -r requirements.txt

Install Spacy's model

spacy download en_core_web_sm

To launch the server:

uvicorn main:app --reload

Frontend

cd frontend

You can use the .nvmrc file to set the correct Node.js version suitable for this project:

nvm use

The first time you run it you need to run:

npm install

To launch the web app:

npm run dev

🐳 Option 2: Docker (dev)

Start both services (frontend + backend) with Docker Compose for a convenient dev environment.

From the root directory:

docker-compose up --build

The compose setup mounts local folders into the containers so code changes are reflected immediately.

🚀 Enjoy it!

• Backend will be available at http://localhost:8000
• Frontend (Vite dev server) at http://localhost:5173

✏️ Contributing

You can contribute to:

• Qux360. Look at the Contribution Guidelines for more details.
• Qux360 Demo. Look at the Contribution Guidelines for more details.

📜 License

Qux360 Demo is licensed under the Apache License 2.0.

IBM ❤️ Open Source AI