Bakery Forecasting Platform - Microservices

Overview

AI-powered demand forecasting platform for bakeries in Madrid, Spain using microservices architecture.

Architecture

• API Gateway: Central entry point for all client requests
• Auth Service: User authentication and authorization
• Training Service: ML model training for demand forecasting
• Forecasting Service: Generate predictions using trained models
• Data Service: External data integration (weather, traffic, events)
• Tenant Service: Multi-tenant management
• Notification Service: Email and WhatsApp notifications

Quick Start

Prerequisites

• Docker and Docker Compose
• Python 3.11+
• Node.js 18+

Setup

# Run setup script (this script!)
./scripts/setup.sh

# Start services
docker-compose up -d

# Check service health
curl http://localhost:8000/health

Services

• Gateway: http://localhost:8000
• API Docs: http://localhost:8000/docs
• Grafana: http://localhost:3002
• Prometheus: http://localhost:9090
• RabbitMQ Management: http://localhost:15672

Development

Running Tests

./scripts/test.sh

Building Services

docker-compose build

Viewing Logs

# All services
docker-compose logs -f

# Specific service
docker-compose logs -f auth-service

Service URLs (Development)

• Gateway: http://localhost:8000
• Auth Service: http://localhost:8001
• Training Service: http://localhost:8002
• Forecasting Service: http://localhost:8003
• Data Service: http://localhost:8004
• Tenant Service: http://localhost:8005
• Notification Service: http://localhost:8006

Environment Variables

Copy .env.example to .env and update the following:

# External API Keys
AEMET_API_KEY=your-aemet-api-key
MADRID_OPENDATA_API_KEY=your-madrid-opendata-key

# Email Configuration
SMTP_USER=your-email@gmail.com
SMTP_PASSWORD=your-email-password

# WhatsApp API
WHATSAPP_API_KEY=your-whatsapp-api-key

# JWT Secret (change in production!)
JWT_SECRET_KEY=your-super-secret-jwt-key-change-in-production

Troubleshooting

Services won't start

# Check if ports are available
docker-compose ps
netstat -tulpn | grep :8000

# Restart services
docker-compose down
docker-compose up -d

Database connection issues

# Check database containers
docker-compose logs auth-db
docker-compose logs training-db

# Reset databases
docker-compose down -v
docker-compose up -d

Service communication issues

# Check service health
curl http://localhost:8000/health
curl http://localhost:8001/health
curl http://localhost:8002/health

# Check RabbitMQ
open http://localhost:15672
# User: bakery, Password: forecast123

Next Steps

1. Configure External APIs: Add your AEMET and Madrid Open Data API keys
2. Test Authentication: Register a user and test login
3. Upload Sales Data: Import historical sales data
4. Train Models: Start your first training job
5. Generate Forecasts: Create demand predictions

License

MIT License