- AWS Solutions Library›
- Guidance for Predictive Scores for Member Retention on AWS
Guidance for Predictive Scores for Member Retention on AWS
Overview
How it works
These technical details feature an architecture diagram to illustrate how to effectively use this solution. The architecture diagram shows the key components and their interactions, providing an overview of the architecture's structure and functionality step-by-step.
Well-Architected Pillars
The architecture diagram above is an example of a Solution created with Well-Architected best practices in mind. To be fully Well-Architected, you should follow as many Well-Architected best practices as possible.
This guidance can be deployed with infrastructure as code and automation for fast iteration and consistent deployments. Use Amazon CloudWatch for application and infrastructure monitoring. Use Amazon SageMaker Model Monitor and Amazon SageMaker Clarify to track bias and model drift.
Use AWS Identity and Access Management (AWS IAM) to ensure users and services have least privilege access, especially to sensitive donor or member data in Amazon S3. Use Amazon Macie to identify possible sensitive data, and obfuscate or remove irrelevant data before using in SageMaker. Use Amazon Virtual Private Cloud (Amazon VPC) to enable connectivity to resources from only the services and users that are needed.
Most services used in the architecture are serverless, and are deployed with high availability by default. Use continuous integration/continuous delivery (CI/CD) practices, and SageMaker Pipelines to automate model development, deployment and management. Collect and automate action on metrics collected in CloudWatch.
Use monitoring to generate alarm-based notifications using CloudWatch, and adjust resources accordingly. Use SageMaker Experiments to optimize algorithms and features.
User SageMaker Studio auto shutdown to avoid paying for unused resources. Start training with small quantities of data. Use Amazon S3 storage classes appropriately, based on data access patterns.
Use managed services when possible, to shift the responsibility of optimizing hardware to AWS. Shut down resources when not in use.
Disclaimer
Did you find what you were looking for today?
Let us know so we can improve the quality of the content on our pages