---
title: Arrakis
emoji: "\U0001F4D6"
colorFrom: indigo
colorTo: purple
sdk: docker
app_port: 7860
pinned: false
license: mit
---

# Visionary Aide

An intelligent AI desktop assistant with chat, document analysis (PDF), voice, and screen-share capabilities. Built with Vite + React + TypeScript, powered by Ollama (local) or OpenAI.

## Features

- **Multi-provider chat** with streaming responses (Ollama or OpenAI)
- **PDF document analysis** — upload, parse, and ask questions
- **Voice input** via Web Speech API
- **Screen share** context for visual queries
- **PWA** installable as a desktop app
- **Local-first** — settings and conversations stored in `localStorage`

## Configuration

Open the in-app **Settings** panel and pick a provider:

- **Ollama** — point at a running Ollama server (default `http://localhost:11434`)
- **OpenAI** — paste your API key and pick a model (default `gpt-4o-mini`)

> Note: In the deployed Space, only **OpenAI** is reachable. Ollama requires a local server, which is not accessible from the public Space.

## Local development

```sh
npm install
npm run dev        # http://localhost:8080
npm run build      # production bundle in dist/
npm run lint
npm run test
```

## Deployment to Hugging Face Spaces

This repo includes a `Dockerfile` and `nginx.conf` for HF Spaces.

1. Create a new Space at <https://huggingface.co/new-space> with **SDK: Docker**.
2. Add HF as a remote and push:
   ```sh
   git remote add hf https://huggingface.co/spaces/<your-username>/<space-name>
   git push hf main
   ```
3. HF will build the image and serve the app on port `7860`.

To log in from the CLI first:

```sh
huggingface-cli login
```

## Tech stack

- Vite 5 + React 18 + TypeScript
- Tailwind CSS + shadcn/ui
- react-router-dom, framer-motion, recharts
- vite-plugin-pwa
- Vitest + Testing Library