[wooders]

I don't think it's simple to deploy scaleable predictions - that's why model hosting solutions like SageMaker's and GCP's AI Platform exist, and there's no need for people to be re-implementing model deployment/monitering.