Machine Learning Model Deployment Serverless Orchestrator

Create a serverless orchestration system for deploying and managing machine learning models using AWS Lambda and Node.js. Build a system that can automatically version, deploy, and route inference requests to appropriate model endpoints. Implement canary deployment strategies, model performance tracking, and automatic scaling based on inference load. Include comprehensive logging and monitoring for model performance and system health.