I have a usecase for hosting multiple xgboost models in one sagemaker endpoint. The models have a slightly different feature set and preprocessing for features.
The t
