MLOps Academy

Learning Paths

Choose a path that matches your background and goals. Follow it linearly, or dip into the steps that feel most relevant right now.

Designed for practical, real-world MLOps work
How to use these paths
Pick one primary path and treat it as a guide. You don't need to complete every step before applying what you learn at work.
Skills-focusedSelf-paced
Data Scientist → MLOps Engineer
Intermediate
Turn notebooks and experiments into production systems.
Current data scientists / ML researchers4–8 weeks (self-paced)
1.

Map your environments

Review Local, Development, Staging and Production environment patterns and map them to your current workflow.

Go to resource
2.

Add real CI to one project

Use the Continuous Integration page to introduce linting, typechecks and tests to an existing ML repo.

Go to resource
3.

Refactor one pipeline for maintainability

Apply code best practices to one core pipeline: separate concerns, configuration, logging and error handling.

Go to resource
4.

Define monitoring and drift signals

Draft what you would monitor for your current model in production: latency, errors, business KPIs and drift.

Software Engineer → MLOps / Platform
Intermediate
Leverage infra and DX skills to build ML platforms.
Backend / infra / platform engineers4–6 weeks (self-paced)
1.

Harden local ML dev

Use the Local Environment track to make local ML dev reproducible: env management, Docker, shell commands.

Go to resource
2.

Experiment with MCP

Build a tiny MCP server from the Development → MCP page and think about how your org could expose tools and resources.

Go to resource
3.

Clarify SLOs and runbooks

From the Production environment mindset, write draft SLOs and a basic incident runbook for one ML service.

Go to resource
Breaking into MLOps
Beginner-friendly
Pair learning with concrete career moves.
Engineers and analysts exploring MLOps roles2–4 weeks (self-paced)
1.

Read an MLOps role overview

Use the Blog and Roles sections to understand how companies describe the MLOps role and its expectations.

Go to resource
2.

Map your skills to job descriptions

Pick one or two roles and explicitly mark where you are strong vs. weak across infra, data, CI/CD, and monitoring.

Go to resource
3.

Connect with practitioners and recruiters

Use Practitioners, Recruiters and Community links from the sidebar to have 2–3 conversations with people doing the work.

Go to resource