In this session we show how engineers, analysts and scientists can develop, train, tune and deploy machine learning models via Amazon SageMaker, a docker-based, framework-agnostic orchestrator dedicated to machine learning with strong primitives for automation, monitoring and deployment. We will cover the following topics:
- Amazon SageMaker architecture and key functionality
- Scaling and deploying open source frameworks on Amazon SageMaker (sklearn, MXNet, Keras, TensorFlow, pyTorch, R)
- Amazon SageMaker built-in Algorithm library: 18 state-of-the art algorithm covering broad use-cases, from computer vision to recommender systems
- ML model deployment