What are the components of a machine learning system?
Overview - Infrastructure and Tooling


  • Google's seminal paper "Machine Learning: The High-Interest Credit Card of Technical Debt" states that if we look at the whole machine learning system, the actual modeling code is very small. There are a lot of other code around it that configure the system, extract the data/features, test the model performance, manage processes/resources, and serve/deploy the model.

  • The data component:

    • Data Storage - How to store the data?

    • Data Workflows - How to process the data?

    • Data Labeling - How to label the data?

    • Data Versioning - How to version the data?

  • The development component:

    • Software Engineering - How to choose the proper engineering tools?

    • Frameworks - How to choose the right deep learning frameworks?

    • Distributed Training - How to train the models in a distributed fashion?

    • Resource Management - How to provision and mange distributed GPUs?

    • Experiment Management - How to manage and store model experiments?

    • Hyper-parameter Tuning - How to tune model hyper-parameters?

  • The deployment component

    • Continuous Integration and Testing - How to not break things as models are updated?

    • Web - How to deploy models to web services?

    • Hardware and Mobile - How to deploy models to embedded and mobile systems?

    • Interchange - How to deploy models across systems?

    • Monitoring - How to monitor model predictions?

  • All-In-One: There are solutions that handle all of these components!