Step 8 - SageMaker

SageMaker

EcommCo’s Data Lake leverages Amazon SageMaker for easy development, training and deployment of machine learning models.

a. Execute SageMaker Notebook

SageMaker Notebooks are fully managed machine learning compute instances running the Jupyter Notebook application. Notebooks allow users to write code to create model training jobs, deploy models into production using Amazon SageMaker hosting service and test or validate produced solutions.

  1. Visit SageMaker Notebook
  2. Review the example code placed in the Jupyter Notebook
  3. Run notebook

b. Review created artifacts

SageMaker uses S3 to store model artifacts, from where they can be deployed into Amazon SageMaker hosting service. After the deployment process, Amazon SageMaker provides an HTTPS endpoint where the machine learning model is available to provide inferences.

  1. Visit AWS Management Console and take a look at:

c. Infer using deployed machine learning model

{% include 'error_box.html' %}

Choose the date to infer sales:

When you click this button, the following steps will be performed within your AWS account:

  • Request for inference to the deployed endpoint is sent