DataMind

⚗️ Your AI-Powered Data Alchemy Cauldron, Turning Documents into Golden Insights

English | 中文

📖 Introduction

Welcome to the magical world of data alchemy! DataMind is an AI-powered knowledge cauldron that unearths hidden insights behind your data.

Data in, Surprise out!

DataMind helps you transform any document into shining crystals of wisdom:

• Put in raw documents, get refined knowledge essence
• Add information fragments, receive coherent complete wisdom
• Insert messy data, obtain clear intuitive insights
• Inject complex problems, extract concise elegant solutions

Whatever challenges you face:

• Need deep market insights? Put in industry reports and news articles — harvest a comprehensive market landscape analysis
• Building complex technical documentation? Add codebase and API docs — get a perfectly structured technical manual
• Analyzing competitor movements? Input competitor data and social information — receive a precise competitive landscape map
• Planning your next project? Place historical files and requirement lists — display a well-organized project blueprint

Like a magic stone turning ordinary into gold, DataMind will complete the magical transformation from ordinary data to shining wisdom:

• Applies cutting-edge AI technology to understand document meanings
• Precisely comprehends your every natural language request
• Automatically generates professional, coherent analysis reports
• Presents final products in your preferred format
• Ensures every output is both logical and factually accurate

Put in your data, state your needs, and let DataMind's AI magic craft the perfect answer!

✨ Core Features

🎨 New Features (v0.3.0)

• First official release version

🚀 Quick Start

Installation

# Clone the project
git clone https://github.com/helixlife-ai/datamind.git
cd datamind

# Install Python dependencies
pip install -r requirements.txt
playwright install chromium

# Generate test data
python scripts/generate_test_data.py

# Install Node.js (if not already installed)
# For Ubuntu/Debian:
curl -fsSL https://deb.nodesource.com/setup_20.x | sudo -E bash -
sudo apt-get install -y nodejs

# For MacOS (using Homebrew):
brew install node@20

# For Windows:
# Please download and install Node.js 20.x version from https://nodejs.org/en/download/

# Install pnpm (if not already installed)
npm install -g pnpm

Usage Example

The easiest way to get started is to run the example script:

# Run the app
cd app
pnpm install
node server.js

Open in browser: http://localhost:3000 Place your local documents in the work_dir directory, then click the start button.

⚙️ Configuration

Environment Variables

The project uses environment variables for configuration. You can set them up in two ways:

1. Using .env file (recommended):

   cp .env.example .env
   # Then edit .env with your values

2. Setting directly in environment with your large language model (Claude3.7 recommended, DEEPSEEK alternative):

   DEFAULT_API_KEY=["your-api-key-here"]
   DEFAULT_BASE_URL=https://api.anthropic.com/v1
   DEFAULT_GENERATOR_MODEL="claude-3-7-sonnet-20250219"
   DEFAULT_REASONING_MODEL="claude-3-7-sonnet-20250219"

🤝 Contributing

PRs and Issues welcome! See Contributing Guide for details.

📄 License

This project is under MIT License - see LICENSE file

👥 Contact

• WeChat: imjszhang
• X: JSZHANG
• Community: DataMind Club
• Email: zhangjingsong@helixlife.cn
• Team: Helixlife AI Lab

🙏 Acknowledgments

Thanks to these open source projects:

• Sentence-Transformers
• FAISS
• DuckDB