Anchor Insight AI

• The application that is taking charge of AI related tasks mainly. 主にAI関連のタスクを担当するアプリケーション
• A unified FastAPI service that integrates focus analysis and scoring capabilities using modular architecture with dependency injection
• We will try to release another version in the future to make sure that this code repository could be used

🏗️ Architecture

This application uses a unified single-service architecture (not microservices):

• Single FastAPI application running on port 7003
• Modular controllers and services with dependency injection
• Direct in-process service calls (no HTTP between services)
• Centralized configuration and routing

🚀 Quick Setup

Prerequisites

• Python 3.9+ (recommended: Python 3.12)
• pipenv (will be installed automatically if not present)
• OpenAI API Key

One-Command Setup

./entry.sh

This will:

• ✅ Check Python version compatibility
• ✅ Install pipenv if needed
• ✅ Create virtual environment
• ✅ Install all dependencies
• ✅ Create .env file with default configuration
• ✅ Verify installation

Configuration

After setup, update your .env file with your OpenAI API key:

OPENAI_API_KEY=your_actual_api_key_here

🎯 Running the Application

Option 1: Using Python Directly

Start the Unified Service (Port 7003)

The application runs as a single unified FastAPI service that integrates focus analysis and scoring capabilities.

pipenv run python src/app/main.py

Or using uvicorn directly:

pipenv run uvicorn src.app.main:app --host 0.0.0.0 --port 7003 --reload

Option 2: Using Docker

# Build and run with docker-compose
cd src
docker-compose up --build

# Or build and run manually
docker build -t anchor-insight-ai -f src/Dockerfile .
docker run -p 7003:7003 --env-file .env anchor-insight-ai

📚 API Documentation

Once the service is running:

• API Documentation (Swagger): http://localhost:7003/docs
• API Documentation (ReDoc): http://localhost:7003/redoc
• Health Check: http://localhost:7003/health
• Root Endpoint: http://localhost:7003/

Available endpoints:

• /api/v1/monitor/* - Focus time monitoring and person detection
• /api/v1/analyze/* - Focus score analysis from images

🛠️ Additional Commands

# Activate virtual environment
pipenv shell

# Run the application
pipenv run python src/app/main.py

# Or using uvicorn with auto-reload for development
pipenv run uvicorn src.app.main:app --host 0.0.0.0 --port 7003 --reload

📁 Project Structure

anchor-insight-AI/
├── entry.sh                      # Setup and dependency installation script
├── Pipfile                       # Python dependencies
├── .env                          # Environment configuration (created from .env.template)
├── src/
│   ├── app/
│   │   └── main.py              # Unified FastAPI application
│   ├── config/
│   │   └── settings.py          # Application settings and configuration
│   ├── controllers/
│   │   ├── focus_controller.py        # Focus monitoring endpoints
│   │   └── focus_score_controller.py  # Focus score analysis endpoints
│   ├── services/
│   │   ├── focus_service.py           # Focus monitoring service
│   │   └── focus_score_service.py     # Focus score analysis service
│   ├── models/
│   │   └── focus_models.py      # Data models and schemas
│   ├── constants/
│   │   └── focus_constants.py   # Application constants
│   ├── dependencies.py           # Dependency injection setup
│   ├── Dockerfile                # Docker container definition
│   └── docker-compose.yml        # Docker orchestration
└── README.md

🔧 Development

Virtual Environment

The project uses pipenv for dependency management:

# Activate virtual environment
pipenv shell

# Install new package
pipenv install package_name

# Install dev dependencies
pipenv install package_name --dev

Environment Variables

Variable	Description	Default
OPENAI_API_KEY	Your OpenAI API key	Required
MODEL_ID	OpenAI model to use	gpt-5-nano
MAX_FILE_SIZE_MB	Maximum file size for uploads	10
URL_TIMEOUT_SECONDS	Request timeout	30
API_HOST	Server host address	0.0.0.0
API_PORT	Server port number	7003
API_RELOAD	Auto-reload on code changes	false
LOG_LEVEL	Logging verbosity	info
CAMERA_INDEX	Camera device index	0
CAMERA_WIDTH	Camera resolution width	640
CAMERA_HEIGHT	Camera resolution height	480
CONFIDENCE_THRESHOLD	YOLO confidence threshold	0.5

🚨 Troubleshooting

Common Issues

Python version error:

# Check Python version
python3 --version
# Ensure you have Python 3.9+

Pipenv not found:

# Install pipenv manually
python3 -m pip install --user pipenv

Service connection errors:

• Ensure the service is running on port 7003
• Check configuration in .env file
• Verify firewall settings allow connections to port 7003

OpenAI API errors:

• Verify your API key in .env file
• Check your OpenAI account balance
• Ensure you have access to the specified model

📝 License

See LICENSE file for details.