AWS Certified Machine Learning Engineer Associate (MLA-C01): The Complete Certification Guide

Important: the MLA-C02 update

AWS is refreshing this certification. As of mid-2026, MLA-C01 is the current version, but registration for the updated exam, MLA-C02, opens on September 1, 2026, and the last day to take MLA-C01 in English is September 28, 2026. If you are scheduling a test date, confirm which version you are booking, and check the official exam guide for the version that applies on your date. The guidance in this post reflects MLA-C01.

What MLA-C01 actually is

MLA-C01 is the AWS Certified Machine Learning Engineer Associate exam. It validates the ability to build, operationalize, deploy, and maintain machine learning solutions and pipelines on AWS. Unlike the foundational AI Practitioner exam, this is a hands-on engineering credential centered on Amazon SageMaker and the surrounding AWS data and MLOps services.

There are no enforced prerequisites, but AWS describes the target candidate as having about one year of experience using Amazon SageMaker and other AWS services for machine learning, plus a background in data engineering and programming. It sits at the associate tier, one step up from foundational.

Exam format at a glance

All figures from the official AWS Certified Machine Learning Engineer Associate (MLA-C01) exam guide. The 15 unscored questions are not identified during the exam and do not affect your result.

The four content domains and their weights

The 50 scored questions are spread across four domains covering the ML lifecycle end to end. AWS uses a compensatory scoring model, so you pass on your overall scaled score, not domain by domain. The weights are close together, so no single domain can be skipped.

Domain 1: Data Preparation for Machine Learning (28%)

The largest domain. It covers ingesting and storing data, transforming and cleaning it, and engineering features for ML. Expect AWS data services (S3, AWS Glue, Amazon EMR, Amazon Athena), SageMaker data tooling such as Data Wrangler and Feature Store, and concepts like data formats, labeling, and handling bias and imbalance in training data.

Domain 2: ML Model Development (26%)

Choosing a modeling approach, training and tuning models, and evaluating them. This covers selecting algorithms and built-in SageMaker models, hyperparameter tuning, training strategies, and evaluation metrics for classification, regression, and other tasks, plus how to interpret model performance and avoid overfitting.

Domain 3: Deployment and Orchestration of ML Workflows (22%)

Getting models into production. It covers deployment options (real-time endpoints, batch transform, serverless and asynchronous inference), choosing compute and scaling, and orchestrating pipelines with SageMaker Pipelines and related AWS services. Infrastructure as code and CI/CD for ML also appear here.

Domain 4: ML Solution Monitoring, Maintenance, and Security (24%)

Keeping deployed ML healthy and secure. It covers monitoring for data and model drift with SageMaker Model Monitor, observability with CloudWatch, cost and performance optimization, and securing ML systems with IAM, encryption, and network controls. This MLOps and security focus is a large part of what makes the exam an engineering credential rather than a knowledge check.

How to prepare

This is a hands-on associate exam, so reading alone is not enough:

• Read the official exam guide first. It lists the task statements and in-scope services for each domain, and it is the definitive scope document, especially with the MLA-C02 update approaching.
• Spend real time in Amazon SageMaker. Building, tuning, deploying, and monitoring a model end to end covers all four domains at once and is the single most effective preparation.
• Balance across domains. The weights range only from 22% to 28%, so unlike some exams there is no small domain you can safely skim. Data preparation and MLOps and security together are more than half the exam.
• Use AWS Skill Builder. AWS publishes an official exam-prep plan and practice questions for MLA-C01.

If you already work with SageMaker day to day, the exam mostly formalizes what you know. If you are coming from a data or software background without much ML deployment experience, give Domains 3 and 4 the most runway.

Is it worth taking?

MLA-C01 is a genuine engineering credential, so it carries more weight with employers than a foundational exam. For data engineers, software engineers, and analysts moving into ML roles, it validates the deploy-and-operate skills that job descriptions increasingly ask for. The main caveat is timing: with MLA-C02 arriving, confirm which version fits your schedule before booking, and weigh whether to sit the current exam now or prepare for the update. As with all AWS certifications, it is valid for three years.

FAQ