👓 Personhood Context

🔍 AI-powered face recognition and person identification system built with Convex

🚀 What it does

This application provides an intelligent pipeline for identifying people in uploaded images:

1. 📸 Image Upload - Accepts images via HTTP endpoint
2. 🤖 Face Recognition - Uses FaceCheck API to find similar faces across the web
3. 🌐 Web Analysis - Analyzes top results using Exa API to extract context
4. 👤 Name Extraction - Identifies human names from relevant web pages
5. 📊 Smart Ranking - Filters and ranks results by relevance score

🛠️ Architecture

HTTP Endpoints

• POST /process - Upload images for face recognition analysis

Core Workflow

Image Upload → Face Recognition → Web Search → Name Extraction → Results

Database Schema

• received - Stores uploaded image metadata
• faceCheck - Maps images to FaceCheck search IDs
• faceCheckUrls - Stores recognition results with confidence scores

🔧 Setup

Environment Variables

EXA_API_KEY=your_exa_api_key
FACECHECK_TOKEN=your_facecheck_token

Dependencies

• FaceCheck API - Face recognition service
• Exa API - Web content analysis and extraction
• Convex - Backend infrastructure

📡 API Usage

Upload an image for analysis:

curl -X POST http://your-convex-site.convex.cloud/process \
  -F "image=@path/to/image.jpg"

🎯 Key Features

• ⚡ Real-time Processing - Instant image analysis pipeline
• 🎯 Smart Filtering - Excludes social media URLs for better results
• 📈 Confidence Scoring - Ranks results by recognition confidence
• 🔒 Secure Storage - Images stored in Convex file storage
• 🌊 Async Processing - Non-blocking background operations

🏗️ Built With

• Convex - Backend framework
• FaceCheck.id - Face recognition API
• Exa - Web search and extraction API
• TypeScript - Type-safe development