Deploy Models using Google's Cloud Machine Learning Engine

[slavakurilyak]

User Story

As a machine learning developer, I want to deploy all predictive or classifications models on Google's Cloud Machine Learning Engine, so that I can deploy on several frameworks, such as:

1. scikit-learn for the breadth and simplicity of classical machine learning
2. XGBoost for the ease and accuracy of extreme gradient boosting
3. Keras for easy and fast prototyping of deep learning
4. TensorFlow for the cutting edge power of deep learning

[slavakurilyak]

Now that Kryptos is running on Google infrastructure (see #61), we can use Google's Cloud Machine Learning Engine (ML Engine) to deploy machine learning models.

Inspiration: Google released support for XGBoost on April 5, 2018. With this release, we can serve machine learning predictions in real time on the Google's Cloud ML Engine. To learn more about this integration, check out the documentation for scikit-learn & XGBoost. Google provides code examples to help developers get started.

@bukosabino this can be an excellent way to serve production-ready machine learning models.