Data Portrait

Data Portrait is a web app that transforms your shopping and reading history into stunning, personalized AI-generated portraits. Connect your accounts from major brands (like Amazon, Wayfair, Office Depot, and Goodreads), and Data Portrait will analyze your purchase data to create unique images that reflect your style, interests, and personality.

Live Demo: https://dataportrait.app/

Features

• Connect Shopping & Reading Accounts: Securely connect your Amazon, Wayfair, Office Depot and Goodreads accounts.
• Automatic Import: Instantly fetch your recent purchases and reading history.
• AI Portrait Generation: Create unique portraits based on your real data, powered by Google Gemini and FLUX.
• Customization: Choose portrait style, gender, and traits for a personalized result.
• Live Data Analysis: Visualize and review the products and brands that shape your portrait.
• Privacy-Focused: Your data is used only for generating your portraits and is never sold or shared.

How It Works

1. Connect Accounts: Use the sidebar to securely link your brand and reading accounts.
2. Import Purchases: The app fetches your order and reading history using the GetGather API.
3. Customize Portrait: Select your preferred style, gender, and traits (e.g., hair, age, features).
4. Generate Portrait: Advanced AI models create a portrait that naturally integrates your interests and purchases.
5. Download & Share: Preview, download, and share your personalized data portrait.

Supported Brands

• Amazon
• Wayfair
• Office Depot
• Goodreads

Technical Overview

• Frontend: React (Vite), TypeScript, Tailwind CSS.
• Backend: Express.js, geolocation via MaxMind, reverse proxy to GetGather API.
• AI Models: Google Gemini via Portkey or direct Google GenAI, or FLUX.
• Storage: Google Cloud Storage (optional) or local filesystem.
• Data Model: Purchases include brand, order date, products, images, etc.

Configuration

Create a .env file in the project root with the following variables:

# GetGather API Configuration
GETGATHER_URL=https://api.getgather.com

# MaxMind GeoIP Configuration (optional)
MAXMIND_ACCOUNT_ID=your_maxmind_account_id
MAXMIND_LICENSE_KEY=your_maxmind_license_key

# AI Providers (required for image generation)
# Image generation provider is automatically selected based on API key availability:
# - If PORTKEY_API_KEY is set, uses Portkey (recommended for production)
# - If only GEMINI_API_KEY is set, uses Google GenAI directly
# - If only FLUX_API_KEY is set, uses FLUX directly
# - Both keys can be set; priority order is: Portkey → Google GenAI → Flux
PORTKEY_API_KEY=your_portkey_api_key
GEMINI_API_KEY=your_gemini_api_key
FLUX_API_KEY=your_flux_api_key

# Google Cloud Storage (optional, for cloud image storage)
# If not configured, images are stored locally in the public/ directory
GCS_BUCKET_NAME=data-portrait-imagegen
GCS_PROJECT_ID=your_gcp_project_id

# Feature Flags
# Set to 'true' to enable face upload feature in the UI
ALLOW_FACE_UPLOAD=false

Development

Using Docker

Use Docker or Podman to pull the container image and run it:

docker run -p 3000:3000 \
  -e GETGATHER_URL=your_local_mcp_getgather_url \
  -e PORTKEY_API_KEY=your_portkey_key \
  ghcr.io/mcp-getgather/data-portrait:latest

Then open localhost:3000 to access the application.

Local Development

npm install
npm run dev