How to make predictions with GCP AI Platform Predictions using HTTP requests

Question

I have deployed a model to AI Platform Predictions successfully by following this tutorial. Now, the standard way I would make predictions with this model is by using the SDK.