Scaling ML from Zero to Millions of Users

Blog series


A small number of machine learning models fail quickly, some look promising and demonstrate a level of predictive power. Then testing and deploying them in a production environment is another challenge where they either fail or prove their worth. This series shall demonstrate how to train and scale up ML models from humble beginnings to world dominations.

AI & Machine Learning
Scaling Machine Learning from 0 to Millions of Users — Part 2 Training: EC2, EMR, ECS, EKS or SageMaker?

Part 1 of this article discussed a few simple techniques that helped with initial scalability of machine learning… and hopefully with reducing manual ops. Since then, despite a few production hiccups due the lack of high availability, life has been pretty good. However, traffic soon starts to increase, data piles up, more models need to be trained, etc. Technical and business stakes are getting higher, and the current architecture will go underwater soon. This post focuses on scaling training to a large number of machines.

11 MINUTES READ Continue Reading »

AI & Machine Learning
Scaling Machine Learning from 0 to Millions of Users — Part 1

So you want to build a ML model. No Machine Learning is easier to manage than no Machine Learning. Figuring a way to use high-level services could save you weeks of work, maybe months. In this series of posts, we’ll discuss how to train ML models and deploy them to production, from humble beginnings to world domination. Along the way, we’ll try to take justified and reasonable steps, fighting the evil forces of over-engineering.

10 MINUTES READ Continue Reading »

  • Top articles, research, podcasts, webinars and more delivered to you monthly.

  • Experfy Content Manager
    Leave a Comment
    Next Post

    Leave a Reply

    Your email address will not be published. Required fields are marked *

    More in Blog series
    Experfy Insights,Blog series
    No Data Engineers, No Problem: How a Small Data Science Team Can Standardize and Centralize Data with a Data Library

    Introduction Access to and control of data is one of the biggest challenges faced by data analysts and data scientists.  Creative, persistent analysts find ways to get access to at least some of this data but doing that efficiently in a way that is also standardized and centralized for everyone on the team is difficult.

    1 MINUTES READ Continue Reading
    Blog series
    AI in Five, Fifty, and Five Hundred Years

    Introduction Prediction is a tricky business. You have to step outside of your comfort zone, your fainted vision of the world and see it thorough across all possible dimensions. In this series, we will discuss the future of “AI”, applications that are yet unexplored.

    1 MINUTES READ Continue Reading
    Blog series
    Ethics of Emerging Technologies

    Introduction: Humans are wired to make tough decisions bringing all the context and principles to bear. Similarly, can devices apply the available information to make the right judgment calls? In this series, we shall discuss some ethical dilemmas faced by emerging technologies.

    1 MINUTES READ Continue Reading