All Articles

Introduction

I just wrote this very brief post to explain what this blog will be about, and why I will write it.

Let’s answer the first question quickly: in this blog I will mostly write about the deployment of ML based solutions.

Now, why ?

As a Data Scientist working in a consultancy firm, I took part in many interesting ML projects. And while all the stages of those projects can be challenging (data collection, ML model development, …), I think the most crucial one is the deployment.

Indeed, a state-of-the-art model is worth nothing if not packaged in a user-friendly solution accessible to the end-user.

In this blog, I will post about how to make the value of AI / ML accessible to the end-user, and so I will try to provide answers to the following questions:

  • How to serve a Machine Learning model ? (On premise / On cloud, API, Microservice, …)
  • How to automate ML solutions’ deployment. (Continous Integration / Continuous Deployment for ML)

machine-learning-in-production.png

Here you are, now you know what type of post you can expect to see in the near future on this blog. Of course I might digress here and there, and post about some other stuff I really want to talk about, but the global line is set.