This is a demonstration of integrating Sagemaker with CircleCI to create an end-to-end ML pipeline, including model training and deployment. It uses a monorepo setup where each model is contained in its own folder. Furthermore, it uses CircleCI's dynamic configs to adapt each pipeline to the model that experienced code changes.
A walkthrough of the project can be found in this Medium post.