datashaper-ai 0.1.0

Local-first data cleaning for CSV files. No cloud uploads, complete privacy.

DataShaper CLI

Local-first data cleaning for CSV files. No cloud uploads, complete privacy.

---

🚀 Quick Start

Installation

pip install datashaper

Basic Usage

# Clean a CSV file
datashaper clean mydata.csv

# Output: mydata_cleaned.csv + quality report

---

✨ Features

• 🔒 100% Local Processing - Your data never leaves your device
• ⚡ Dual-Engine Architecture - Polars (fast) + DuckDB (stable)
• 🧹 Auto-Detection - Finds duplicates, nulls, formatting issues
• 📊 Quality Scoring - Get a data quality score (0-100%)
• 🎯 Smart Routing - Automatically selects best engine for your file
• 💼 Enterprise Ready - Hardware-bound licensing, Team features

---

📋 Commands

1. Clean Data

datashaper clean input.csv

What it does:

• Removes duplicate rows
• Handles missing values
• Trims whitespace
• Validates data types
• Outputs: input_cleaned.csv

Example Output:

🔑 License: PRO tier
🧹 Cleaning: customer_data.csv
✅ Cleaned data saved to: customer_data_cleaned.csv

📊 Summary:
  • Issues found: 847
  • Fixes applied: 847
  • Quality score: 94.2%
  • Engine used: Polars

2. Create Bundle (Pro+)

datashaper bundle input.csv -o output.zip

Creates ZIP with:

• cleaned.csv - Cleaned data
• report.html - Detailed quality report
• schema.json - Data schema

3. Activate License

datashaper activate DS-ABCD-1234-WXYZ-5678

Features:

• Binds to your hardware (prevents sharing)
• Works offline after activation
• Stored in ~/.datashaper/license.json

4. Check Status

datashaper status

Shows:

• Current tier (FREE, PRO, TEAM, ENTERPRISE)
• File size limit
• Feature availability
• License expiry (if applicable)

---

🎯 Tier Comparison

Feature	Free	Pro	Team	Enterprise
File Size	25 MB	1 GB	5 GB	Unlimited*
Engine	Polars	Both	Both	Both + Chunked
ZIP Export	❌	✅	✅	✅
HTML Reports	❌	✅	✅	✅
Team Sharing	❌	❌	✅ (10 seats)	✅ (Custom)
Priority Support	❌	❌	✅	✅
Offline Activation	✅	✅	✅	✅

*Unlimited with recommended hardware (see HARDWARE.md)

---

🏗️ Architecture

CrashGuard Engine

Intelligent engine selection:

File < 1GB → Polars (fast, in-memory)
File > 1GB → DuckDB (stable, out-of-core)
Polars OOM → Auto-retry with DuckDB

Why two engines?

• Polars: 10x faster for typical files
• DuckDB: Handles files larger than RAM
• CrashGuard: Never crashes, always falls back

---

🔒 Privacy & Security

Your Data Never Leaves Your Device

Zero cloud uploads:

• All processing happens locally
• No servers see your data
• Works completely offline (after activation)

Network requests (exactly 3):

1. License validation (license key only)
2. Auto-update check (version number only)
3. Team sync (Team tier only, shared rules)

Read more:

• SECURITY.md - Complete security overview
• PRIVACY.md - Privacy policy
• NETWORK_AUDIT.md - Complete network transparency

---

💻 Desktop App

Cross-platform desktop app available:

• Windows (.exe)
• macOS (.dmg)
• Linux (.AppImage)

Features:

• Drag-and-drop file upload
• Real-time progress
• Visual results dashboard
• Team collaboration (Team tier)
• Auto-updates

Download Desktop App →

---

📦 Enterprise Features

Team Tier

Collaboration features:

• 10 user seats
• Shared cleaning rules
• Team activity audit logs
• Centralized billing

Enterprise Tier

For large organizations:

• Unlimited seats
• Custom file size limits
• On-premises deployment
• SSO integration
• Dedicated support
• SLA guarantees

Contact Sales →

---

🛠️ Development

From Source

git clone https://github.com/yourusername/datashaper.git
cd datashaper
pip install -e .

Run Tests

pytest tests/

Build Desktop App

cd desktop
npm install
npm run tauri:build

---

📚 Documentation

• Security & Privacy
• Hardware Recommendations
• Network Audit
• Privacy Policy

---

🤝 Support

Community

• GitHub Issues
• Discussions

Enterprise

• Email: [e-mail]
• Support: [e-mail]
• Security: [e-mail]

---

📄 License

MIT License - see LICENSE for details

Commercial licensing available for Enterprise tier.

---

🚀 Getting Started

1. Install: pip install datashaper
2. Activate: datashaper activate YOUR-LICENSE-KEY
3. Clean: datashaper clean yourfile.csv

That's it! Your data is cleaned locally, privately, securely.

---

Made with ❤️ for data professionals who value privacy.