HeritageAI

AI-powered platform for generating, analyzing, and exporting culturally-inspired design assets
Patterns, palettes, briefs, and more — for any culture, powered by Google Gemini & CLIP

Features • Usage • API • Extending • Dependencies

---

🚀 Overview

HeritageAI is a modern toolkit for designers, developers, and creators to generate authentic, culturally-inspired design assets using advanced AI models. Instantly create patterns, color palettes, design briefs, and more for any culture — with both a beautiful web UI and a powerful CLI.

---

🗂️ Project Structure

heritage-ai/
│
├── mvp_ai/
│   ├── main.py                  # CLI and orchestration module
│   ├── ai_culture_generator.py  # AI-powered culture-specific design analysis
│   ├── web_api.py               # Flask web API and UI
│   ├── clip_model.py            # CLIP-based image scoring
│   ├── gemini_client.py         # Gemini API client
│   ├── imagegen_gemini.py       # Pattern image generation using Gemini
│   ├── color_palette.py         # Color palette extraction utilities
│   ├── export_formats.py        # Exporting assets in various formats
│   ├── gemini_notes.py          # Cultural note generation
│   ├── assets/                  # Generated images, palettes, notes, kits
│   └── ... (tests, docs)
└── README.md

---

✨ Features

Feature	Description
🎨 Pattern Generation	Generate culturally-inspired pattern images using AI
🌈 Color Palette Extraction	Extract color palettes from generated or uploaded images
🤖 AI-Powered Analysis	Analyze images for fonts, motifs, colors, patterns, and design briefs
📦 Bundling & Export	Bundle images, palettes, and notes; export in SVG, PNG, CSS, JSON, Figma...
🖥️ Web UI & REST API	Modern Flask web interface and RESTful API
🛠️ CLI Tools	Command-line tools for batch processing, analysis, and export
🔄 Multi-API Key Support	Robust retry logic with automatic Gemini API key rotation
📝 Professional Codebase	Type hints, docstrings, and logging throughout

---

🏗️ Main Components

• main.py — CLI for all major operations (generation, analysis, export)
• ai_culture_generator.py — AI-powered image analysis for culture metadata
• web_api.py — Flask web UI & API endpoints
• Supporting modules:
  • clip_model.py, gemini_client.py, imagegen_gemini.py, color_palette.py, export_formats.py, gemini_notes.py

---

💻 Usage

CLI (Command Line)

# Generate pattern images
python -m mvp_ai.main generate --culture yoruba --count 3 --aspect-ratio 1:1

# Generate a complete design kit
python -m mvp_ai.main generate-kit --culture edo

# Extract color palettes
python -m mvp_ai.main palette

# Generate cultural briefs
python -m mvp_ai.main brief

# Score images with CLIP
python -m mvp_ai.main clip-score --prompt "Yoruba textile pattern"

# Generate AI-powered culture metadata from an image
python -m mvp_ai.main generate-culture-metadata --culture maori --image path/to/image.png

🌐 Web UI

1. Start the server:

   python -m mvp_ai.web_api

2. Open your browser: http://localhost:5000
3. Generate & download kits for any culture with a click!

---

🔗 API Endpoints

Endpoint	Method	Description
/api/generate-kit	POST	Generate a design kit for a culture
/api/download-kit/<culture>	GET	Download a zipped kit
/api/cultures	GET	List available cultures
/api/clip-score	POST	Score images with CLIP
/api/generate-culture-metadata	POST	Generate metadata for a culture/image

---

🧩 Extending the Project

• Add new cultures by using the UI or CLI/API with a new culture name
• Integrate other AI models by extending gemini_client.py
• Add new export formats in export_formats.py

---

⚙️ Dependencies

• Python 3.x
• Flask
• Gemini API (Google)
• CLIP model
• Pillow, ColorThief (for image and color processing)
• See mvp_ai/requirements.txt for full list

---

📁 Assets

All generated images, palettes, notes, and kits are stored in:

mvp_ai/assets/

---

---

Made with ❤️ for cultural creativity