Briefly, what's SageMaker?

Today I decided to start a new series of short posts that give a high-level basic understanding of some piece of technology. The first post is about AWS SageMaker!

SageMaker is part of the Machine Learning solutions provided by Amazon. These solutions are provided in stages:

Steps for ML using AWS:

  1. Pre-built services to solve standard problems -> Application services (Amazon Rekognition, Polly, Comprehend)
  2. Basic customization (of datasets and algorithms): SageMaker
  3. Full control (including infrastructure): EC2 and Deep Learning AMI.

To use SageMaker, we only need to write instrumenting code, like tune algorithms and declare where the data is stored.