🚀 AI Closet Scanner

Your Personal Fashion AI Assistant

Scan your closet with AI • Get outfit recommendations • Virtual try-on • 100% Private & Encrypted

---

📋 Table of Contents

• Overview
• Features
• Technology Stack
• Architecture
• Getting Started
• Backend Setup
• iOS App Setup
• Daytona Deployment
• API Documentation
• Privacy & Security
• Demo
• License

---

🎯 Overview

AI Closet Scanner is a privacy-first wardrobe management app that uses cutting-edge AI to help you:

• 📸 Scan & Analyze clothing items with computer vision
• 👔 Generate Outfits based on occasion, weather, and style
• 🎨 Virtual Try-On to see how outfits look before wearing them
• 🗣️ Voice Recommendations with natural-sounding style advice
• ☁️ Cloud Backup with end-to-end encryption

The Problem

"I have nothing to wear!" - Despite having a full closet, we often struggle to:

• Remember what we own
• Create cohesive outfits
• Match items appropriately
• Dress for the occasion

Our Solution

AI-powered wardrobe management that:

• ✅ Automatically catalogs your closet
• ✅ Suggests outfits based on context
• ✅ Shows you how outfits look before getting dressed
• ✅ Keeps all your data private and encrypted

---

✨ Features

Core Features

1. AI Clothing Analysis (Claude)

• Upload photos of clothing items
• Get detailed analysis: type, color, pattern, style, seasonality
• Automatic pairing suggestions
• Material and care instructions

2. Smart Outfit Generation (Claude)

• Context-aware outfit suggestions
• Weather integration
• Occasion-based recommendations
• Color harmony analysis
• Alternative suggestions

3. Virtual Try-On (Nano Banana)

• See how outfits look on you before wearing them
• Photorealistic visualization
• Multiple item combinations
• Background removal for clean images

4. Voice Recommendations (ElevenLabs)

• Natural-sounding voice guidance
• Personalized styling tips
• Multi-language support
• Downloadable audio recommendations

5. Privacy-First Storage

• Local encryption (AES-256-GCM)
• Keys stored in iOS Keychain
• Optional cloud backup to Tigris
• No tracking, no data selling

6. Cloud Development (Daytona)

• Instant development environment
• Zero DevOps setup
• Live deployment
• Collaborative workspace

---

🛠️ Technology Stack

AI Services

Service	Purpose	Integration
🤖 Claude (Anthropic)	Fashion analysis & outfit generation	Vision API + Text generation
🗣️ ElevenLabs	Voice recommendations	Text-to-speech
🍌 Nano Banana (Gemini)	Virtual try-on image generation	Gemini 2.0 Flash
☁️ Tigris	Encrypted cloud storage	S3-compatible storage
💳 Brex	Premium subscriptions	Payment processing
📊 Galileo	LLM observability	Request tracing
🏃 Daytona	Cloud development platform	Workspace hosting

Backend (Python)

• FastAPI - Modern async web framework
• Anthropic SDK - Claude API integration
• ElevenLabs SDK - TTS integration
• Google Generative AI - Gemini integration
• Boto3 - S3/Tigris storage
• Galileo Observe - LLM monitoring

Frontend (iOS)

• Swift - Native iOS development
• SwiftUI - Modern declarative UI
• CryptoKit - AES-256-GCM encryption
• Core Data - Local data persistence
• Keychain - Secure key storage

---

🏗️ Architecture

┌─────────────────────────────────────────────────────────┐
│                  iOS APP (Swift/SwiftUI)                 │
│  • Camera capture                                        │
│  • Local encryption (AES-256-GCM)                       │
│  • Core Data storage                                     │
│  • Calls backend API                                     │
└────────────────┬────────────────────────────────────────┘
                 │
                 │ HTTPS
                 ▼
┌─────────────────────────────────────────────────────────┐
│          BACKEND API (Python/FastAPI)                    │
│          🏃 HOSTED ON DAYTONA                            │
│                                                           │
│  ┌────────────────────────────────────────────────────┐ │
│  │  AI Service Integrations:                          │ │
│  │  • Claude API (Anthropic)                          │ │
│  │  • ElevenLabs API                                  │ │
│  │  • Nano Banana (Gemini API)                        │ │
│  │  • Tigris S3 Storage                               │ │
│  │  • Galileo LLM Tracing                             │ │
│  └────────────────────────────────────────────────────┘ │
└─────────────────────────────────────────────────────────┘

Why This Architecture?

iOS App (Local):

• Privacy-first: data encrypted locally
• Offline functionality
• Fast, responsive UI
• Native iOS experience

Backend on Daytona:

• Heavy AI processing in the cloud
• API aggregation (one backend, multiple AI services)
• Monitoring with Galileo
• Easy to demo and scale

---

🚀 Getting Started

Prerequisites

• Backend:

  • Python 3.11+
  • Daytona account (for deployment)
  • API keys (see below)

• iOS:

  • macOS with Xcode 15+
  • iOS 17+ simulator or device
  • Apple Developer account (for device testing)

Required API Keys

You'll need API keys from the following services:

1. Anthropic Claude: https://console.anthropic.com/
2. ElevenLabs: https://elevenlabs.io/
3. Google Gemini: https://makersuite.google.com/app/apikey
4. Tigris: https://console.tigris.dev/
5. Galileo: https://console.galileo.ai/

Optional: 6. Brex: https://developer.brex.com/

---

🔧 Backend Setup

1. Clone the Repository

git clone https://github.com/your-username/closet-scanner.git
cd closet-scanner/backend-api

2. Create Virtual Environment

python3 -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

3. Install Dependencies

pip install -r requirements.txt

4. Configure Environment Variables

cp .env.example .env

Edit .env and add your API keys:

# Anthropic (Claude)
ANTHROPIC_API_KEY=sk-ant-api03-your-key-here

# ElevenLabs
ELEVENLABS_API_KEY=your_elevenlabs_key

# Google Gemini
GEMINI_API_KEY=your_gemini_key

# Tigris
TIGRIS_ACCESS_KEY=tid_your_access_key
TIGRIS_SECRET_KEY=tsec_your_secret_key

# Galileo
GALILEO_API_KEY=your_galileo_key

# Application
BACKEND_URL=http://localhost:8000

5. Run Locally

cd backend-api
uvicorn app.main:app --reload

The API will be available at: http://localhost:8000

• API Docs: http://localhost:8000/docs
• Health Check: http://localhost:8000/health
• Dashboard: http://localhost:8000/dashboard

---

📱 iOS App Setup

1. Open Xcode Project

open ClosetAI.xcodeproj

2. Update Backend URL

Edit ClosetAI/Services/APIClient.swift:

// Change this to your Daytona backend URL
private let baseURL = "https://your-daytona-url.daytona.app"

3. Build and Run

• Select a simulator or device
• Press Cmd + R to build and run

---

🏃 Daytona Deployment

1. Install Daytona CLI

curl -sf https://download.daytona.io/daytona/install.sh | sh

2. Login to Daytona

daytona login

3. Create Workspace

daytona create --name closet-scanner-backend \
  --from-git https://github.com/your-username/closet-scanner

4. Set Environment Variables

In the Daytona dashboard, add your API keys as secrets.

5. Start the Backend

daytona code closet-scanner-backend

Inside the Daytona workspace:

cd backend-api
pip install -r requirements.txt
uvicorn app.main:app --host 0.0.0.0 --port 8000

6. Get Public URL

daytona url closet-scanner-backend

This will return your public URL: https://closet-scanner-backend-xxx.daytona.app

Update the iOS app's APIClient.swift with this URL.

---

📚 API Documentation

Endpoints

Health Check

GET /
GET /health

Clothing Analysis

POST /analyze-clothing
Content-Type: multipart/form-data

Request:
- file: Image file (JPEG, PNG)

Response:
{
  "type": "shirt",
  "color": "blue",
  "pattern": "solid",
  "style": "casual",
  "season": ["spring", "summer"],
  "pairs_well_with": ["jeans", "chinos"],
  "confidence": 0.95
}

Outfit Generation

POST /generate-outfit
Content-Type: application/json

Request:
{
  "wardrobe_items": [...],
  "occasion": "work",
  "weather": {"temperature": 72, "condition": "sunny"}
}

Response:
{
  "items": [...],
  "reasoning": "...",
  "style_tips": "...",
  "audio_url": "https://..."
}

Virtual Try-On

POST /virtual-tryon
Content-Type: application/json

Request:
{
  "user_image_base64": "...",
  "clothing_items_base64": [...]
}

Response: PNG image

For complete API documentation, visit: http://your-backend-url/docs

---

🔒 Privacy & Security

Local Encryption

All user data is encrypted using AES-256-GCM:

• Encryption keys stored securely in iOS Keychain
• Data encrypted before leaving the device
• Only user has access to decryption keys

Cloud Backup (Optional)

If user enables cloud backup:

• Data is already encrypted before upload
• Stored in Tigris S3-compatible storage
• No one (including us) can decrypt without the key

No Tracking

• No analytics
• No user profiling
• No data selling
• No third-party tracking

---

🎥 Demo

Screenshots

[Add screenshots here]

Video Demo

[Add video demo link here]

Live Dashboard

Visit the live metrics dashboard: http://your-backend-url/dashboard

---

📄 License

MIT License - see LICENSE file for details

---

🙏 Acknowledgments

Built with amazing technologies from:

• Anthropic (Claude AI)
• ElevenLabs (Voice AI)
• Google (Gemini AI)
• Tigris (Cloud Storage)
• Galileo (LLM Observability)
• Daytona (Cloud Development)
• Brex (Payments)

---

📧 Contact

Have questions? Found a bug? Want to contribute?

• GitHub Issues: github.com/your-username/closet-scanner/issues
• Email: your-email@example.com

---

Made with ❤️ for the Hackathon

Empowering personal style with AI, while respecting privacy.