Probly

An AI-powered spreadsheet application that combines spreadsheet functionality with Python data analysis capabilities.

Features

• Interactive Spreadsheet: Full-featured spreadsheet with formula support
• Python Analysis: Run Python code directly in your browser using WebAssembly
• Data Visualization: Create charts and visualizations from your data
• AI-Powered: Get intelligent suggestions and automated analysis
• Prompt Library: Access predefined prompts or save your own for quick reuse

Architecture

Probly uses a modern architecture:

• Frontend: Next.js application that runs in the browser
• Python Execution: Pyodide (Python compiled to WebAssembly) runs entirely in the browser
• LLM Integration: OpenAI API calls are proxied through the server

This design means that data analysis happens locally in your browser, providing better performance and privacy.

Requirements

• Node.js 18 or higher
• npm or yarn
• A modern web browser (Chrome, Firefox, Edge, or Safari)
• OpenAI API key

Quick Start with Docker

The easiest way to get started with Probly is using Docker:

1. Clone the repository:

   git clone https://github.com/PragmaticMachineLearning/probly.git
   cd probly

2. Create a simple .env file with your OpenAI API key:

   OPENAI_API_KEY=your_api_key_here
   

3. Build and start the application:

   docker compose build
   docker compose up -d

4. Access Probly at http://localhost:3000

The Docker container serves the Next.js application and handles API requests to OpenAI, while all Python code execution happens in your browser using WebAssembly.

Manual Installation

If you prefer to run Probly without Docker:

1. Clone the repository

   git clone https://github.com/PragmaticMachineLearning/probly.git
   cd probly

2. Install dependencies:

   npm install

3. Create a .env file with your OpenAI API key:

   OPENAI_API_KEY=your_api_key_here
   

4. Start the development server:

   npm run dev

5. For production, build and start:

   npm run build
   npm start

Using Probly

1. Start the application and open it in your browser
2. Import data using the import button or start with a blank spreadsheet
3. Open the AI chat with the keyboard shortcut:
   • Windows/Linux: Ctrl+Shift+?
   • Mac: ⌘+Shift+? (Command+Shift+?)
4. Access the prompt library to use predefined prompts or save your own
5. Ask questions about your data or request analysis

Keyboard Shortcuts

Action	Windows/Linux	Mac
Toggle AI Chat	Ctrl+Shift+?	⌘+Shift+? (Command+Shift+?)
Open Prompt Library	Ctrl+Shift+L	⌘+Shift+L (Command+Shift+L)

Advanced Docker Options

Alternative Docker Setup

You can also pass the API key directly to Docker Compose:

OPENAI_API_KEY=your_api_key_here docker compose build
OPENAI_API_KEY=your_api_key_here docker compose up -d

Building Custom Docker Images

To build a custom Docker image with your API key baked in:

docker build -t probly:custom --build-arg OPENAI_API_KEY=your_api_key_here .
docker run -p 3000:3000 probly:custom

Tech Stack

• Frontend: Next.js 14, TypeScript, React
• Spreadsheet: Handsontable, HyperFormula
• Python Runtime: Pyodide (WebAssembly)
• LLM: OpenAI API
• Visualization: ECharts

Documentation

For comprehensive documentation, visit the Probly Documentation.

License

MIT License