MLA-C02

MLA-C02 exam overview

The AWS Certified Machine Learning Engineer - Associate MLA-C02 exam validates your ability to build, operationalize, deploy, and maintain AI and machine learning solutions on AWS.

The updated exam covers traditional machine learning alongside generative AI, foundation models, retrieval-augmented generation, and agentic workflows. Amazon SageMaker AI remains central to the exam, while Amazon Bedrock now has a much larger role.

MLA-C02 is an engineering and operations exam. It tests how candidates prepare data, develop models, select deployment infrastructure, automate workflows, monitor production systems, control costs, and secure AI and ML resources.

Page contents

1. Current Exam Status

2. Exam Quick View

3. Target Audience and Experience

4. Tasks Validated

5. Exam Domains and Weights

6. Skills and Technologies Covered

7. Preparation Priorities

8. Common Preparation Mistakes

9. CertQueen Preparation Materials

10. Disclaimer

Current MLA-C02 exam status

AWS released the MLA-C02 exam guide on September 1, 2026. The updated exam is currently entering its beta period.

The official exam guide is named MLA-C02. During the beta period, candidates schedule the exam under the code ME1-C02. This beta scheduling code should not be confused with the exam guide version.

MLA-C02 exam quick view

The beta includes additional questions for statistical evaluation. These extra questions account for the longer appointment. AWS does not identify which questions are used for evaluation.

Target audience and recommended experience

MLA-C02 is intended for professionals who build and operate traditional ML and generative AI solutions in production environments.

• Machine learning engineers
• MLOps engineers
• LLMOps engineers
• Data engineers
• Backend software developers
• DevOps engineers
• Data scientists moving into engineering roles

Recommended experience

AWS recommends at least one year of experience using Amazon SageMaker AI, Amazon Bedrock, and other AWS services for ML engineering.

Candidates should also have at least one year of experience in a related role, such as backend development, DevOps, data engineering, or data science. Experience with both traditional machine learning and generative AI is expected.

Knowledge expected before the exam

• Common ML algorithms and their use cases
• Foundation model capabilities and limitations
• Data formats, ingestion, transformation, and feature preparation
• Software engineering and debugging practices
• CI/CD pipelines and infrastructure as code
• Provisioning and monitoring cloud ML resources
• AWS identity, encryption, and data protection controls

Tasks validated by MLA-C02

The exam measures your ability to:

• Ingest, transform, validate, and prepare data for AI and ML modeling.
• Select a modeling approach and train traditional or generative AI models.
• Tune hyperparameters and evaluate model performance.
• Manage model versions and deployment artifacts.
• Select endpoints, compute resources, and scaling configurations.
• Automate AI and ML workflows through CI/CD pipelines.
• Build and operate agentic workflows.
• Monitor models, agents, data, and infrastructure.
• Optimize workload efficiency and cost.
• Secure AI and ML systems through access controls, encryption, compliance features, and AWS best practices.

MLA-C02 exam domains and weights

The percentages apply to scored content. Data Preparation is the largest domain, but the remaining three domains each account for almost one quarter of the scored exam.

Skills and technologies covered

• Amazon SageMaker AI
• Amazon Bedrock and foundation models
• Retrieval-augmented generation
• Agentic AI workflows
• Data ingestion and feature engineering
• Model training, tuning, and evaluation
• Model registries and version management
• Real-time and batch deployment
• CI/CD, MLOps, LLMOps, and infrastructure as code
• Monitoring, observability, drift, and cost control
• AWS identity, encryption, and compliance controls

Preparation priorities

Begin with data preparation because it carries the highest weight and supports every later stage of the ML lifecycle.

1. Practice selecting storage, ingestion, transformation, and feature engineering services.
2. Compare traditional models, foundation models, fine-tuning, prompt engineering, and RAG.
3. Review SageMaker AI and Amazon Bedrock deployment options.
4. Build CI/CD workflows for training, evaluation, registration, approval, and deployment.
5. Study agent orchestration, observability, guardrails, and failure handling.
6. Practice detecting drift, data quality problems, endpoint errors, and excessive cost.
7. Review IAM permissions, encryption, network isolation, audit logging, and compliance controls.

Scenario questions often contain more than one workable option. Choose the answer that meets the stated requirements with the least unnecessary operational work.

Common preparation mistakes

Preparing only for traditional machine learning

MLA-C02 adds foundation models, Amazon Bedrock, RAG, generative AI, and agentic workflows. MLA-C01 materials alone do not cover the complete updated scope.

Treating Amazon Bedrock like SageMaker AI

The two services overlap in some use cases, but they provide different model access, customization, training, deployment, and operational workflows. Practice choosing between them from a set of requirements.

Skipping data validation

Many candidates focus on model training and forget schema changes, missing values, class imbalance, leakage, labeling quality, and other data problems that affect production models.

Choosing the most complex deployment

A real-time endpoint is not always necessary. Review when to use real-time inference, serverless inference, asynchronous inference, batch transform, provisioned throughput, or another deployment option.

Monitoring only endpoint availability

A healthy endpoint can still serve a poor model. Review data quality, drift, model performance, agent traces, latency, errors, throughput, resource use, and cost.

Leaving security until the end

Security appears throughout the ML lifecycle. IAM, encryption, network controls, data access, model access, logging, and compliance should be considered during design and deployment, not added after release.

CertQueen MLA-C02 preparation materials

CertQueen MLA-C02 preparation materials are organized around the four official content domains. They cover both traditional ML engineering and the generative AI additions introduced in the updated exam.

What is included?

• Questions covering all four official domains
• Amazon SageMaker AI and Amazon Bedrock scenarios
• Traditional ML, foundation model, RAG, and agentic AI coverage
• Deployment, CI/CD, monitoring, cost, and security questions
• Answers with concise explanations
• PDF and testing-engine formats
• 12 months of free content updates, with an optional upgrade to 24 months for $10

Start with an untimed session and group missed questions by domain. Rebuild weak workflows in an AWS lab, then move to timed practice after you can explain why the incorrect options fail.

The standard package includes free MLA-C02 content updates for 12 months from the purchase date. Updates may include corrections and changes required when AWS moves the exam from beta to general availability.

 Start Your MLA-C02 Exam Preparation

Disclaimer

CertQueen is an independent preparation provider and is not affiliated with or endorsed by Amazon Web Services. AWS product names and trademarks belong to their respective owners. Preparation materials do not guarantee a passing result.