MLOps — How to start this journey?

In this post, I will discuss step-by-step guidance to learn MLOps. But before we start, few important points or cautions

  • MLOps is not about throwing a tool.
  • MLOps is an Engineering discipline like DevOps or SDLC.
  • Mastering MLOps needs continuous learning and it will only come by spending sufficient time in various kinds of project implementation, reading various patterns, architectures, and so on.
  • MLOps is a multi-disciplinary area that includes knowledge of AI/ML, DevOps, Data Engineering, Cloud Native patterns, Infrastructure, Version Control, ethics, etc.

So, the moral is, there is no shortcut or a quicker way to master. But we can always take a curated and methodical step to start the journey.

So, let's get started…

Pre-Requisite

  • Basic training on classical Machine Learning.
  • Basic training on either Pytorch or Tensorflow framework to understand the code construct.
  • Knowledge of Pandas, Numpy, SQL
  • Python

Step #1: A High-level Overview

A quick high level read on the basics of MLOps with Architecture in this series (MLOps aka Operational AI)

Duration: 30 mins

Step #2: MLOps for beginners — Objective to Production

I found this series quite good and covers E2E of MLOps.

Why?

  • The 2nd top Github repo in MLOps
  • Videos and hands-on with Notebooks to practice
  • Very lucid and easy to progress tutorials
  • Not specific to any particular cloud products

Duration: 3–4 weeks

Step #3: Understand related concepts

This may be in parallel with Step #2 as individuals are going through the chapters. Please note that below are many curated lists. But please feel free to explore more.

Duration: 1 week

  • What is Cloud Native and why we need it?
  • Core Principals of Cloud Native architecture.
  • Adoption of Cloud-Native Architecture, Part 1: Architecture Evolution and Maturity
  • Understand what is CNCF landscape
  • Read about Microservices architecture and its relevance and challenges

Step #4: Explore concepts and technologies more deeper

This step is going into more detail and getting yourself familiar with cloud-native tools and technologies and concepts worth understanding it. It is not a complete list as this area is extremely vast and diverse.

Duration: 4–6 weeks (May change depends on the depth you are planning to go)

  1. Docker Tutorial for Beginners

2. Kubernetes for container orchestration

  • Complete Kubernetes Tutorial for Beginners
  • Understanding Kubernetes networking basics

3. Understand API Gateway and Service Mesh concept

  • Difference between API Gateway and Service Mesh
  • Istio Service Mesh Explained

4. Ingress

  • Tales of the Kubernetes ingress networking

5. CI/CD tools like Jenkins-X, Argo CD, Flux CD

6. GitOps and Kubernetes CD

  • GitOps and Kubernetes: An overview
  • Tutorial: Everything You Need To Become a GitOps Ninja — Alex Collins & Alexander Matyushentsev

7. Seldon Core for deploying ML Models and Monitoring

  • Seldon Core installation on Kubernetes (Openshift)
  • Seamless MLOps with Seldon and MLflow

8. Patterns

  • Sidecar Pattern
  • Microservices Design Pattern
  • What are microservices really all about? — Microservices Basics Tutorial

The list is endless, but I will stop here. These steps and links will give quick boost to learn MLOps and all associated concepts faster.

Happy Learning!

Appendix

--

--

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
AI & Data Engineering

AI & Data Engineering

A Data Enthusiast, Lead Architect with 17 yrs experience in the field of AI Engineering, BI, Data Warehousing, Dimensional modeling, ML and Big Data