AWS Certified Machine Learning Engineer – Associate (MLA-C01) Guide: Legacy Exam Outline
A legacy MLA-C01 guide covering the four exam domains. MLA-C02 beta is available; confirm exam code, language, and version before preparing.
Version notice (2026-10-02): This article covers the legacy MLA-C01 exam. MLA-C02 beta delivery began on 2026-09-29; the beta is currently in English, and AWS lists its exam code as ME1-C02. The English MLA-C01 exam ended on 2026-09-28. Korean, Japanese, and Simplified Chinese MLA-C01 remain available until MLA-C02 general availability. Check the current AWS exam information before booking or studying. Do not use this article to prepare for MLA-C02.
The legacy AWS Certified Machine Learning Engineer – Associate (MLA-C01) outline covers data preparation, model development, workflow deployment, and ML solution monitoring on AWS. It focuses on engineering delivery, not pure algorithm research.
Who is the legacy MLA-C01 exam for?
The exam suits engineers responsible for ML training, deployment, pipelines, or monitoring, and data engineers who bring models into production. Basic ML, cloud IAM, and networking knowledge helps; experience with SageMaker training, endpoints, or pipelines is useful.
AIF-C01 covers AI/ML concepts; DEA-C01 focuses on data pipelines; this exam focuses on the ML engineering lifecycle from preparation to operations.
Legacy MLA-C01 exam domains
These weights apply to the legacy MLA-C01 scored content only. They do not apply to MLA-C02.
| Domain | Official name | Main focus | Weight |
|---|---|---|---|
| 1 | Data Preparation for Machine Learning (ML) | Ingestion, transformation, feature engineering, and data integrity | 28% |
| 2 | ML Model Development | Model selection, training, tuning, and evaluation | 26% |
| 3 | Deployment and Orchestration of ML Workflows | Deployment, inference endpoints, and workflow orchestration | 22% |
| 4 | ML Solution Monitoring, Maintenance, and Security | Drift, maintenance, monitoring, and security | 24% |
Domain 1: Data Preparation for Machine Learning
Study data ingestion, storage, transformation, feature engineering, and data integrity. Check whether training, validation, and inference use consistent data.
Common mix-up: missing feature leakage, class imbalance, or train/serve skew. Define the purpose of each dataset and the checks it needs.
Domain 2: ML Model Development
Learn to select models and metrics for a problem, train and tune models, and analyze performance, cost, and failure modes.
Common mix-up: choosing a metric that does not match the business goal or missing overfitting. Define the problem type and success criteria before comparing results.
Domain 3: Deployment and Orchestration of ML Workflows
Understand when to use real-time, batch, or asynchronous inference, and how to deploy ML workflows with pipelines and CI/CD.
Common mix-up: treating a working endpoint as a complete deployment while ignoring model versions, automated tests, scaling, and rollback. Map the full flow: train, register, deploy, and validate.
Domain 4: ML Solution Monitoring, Maintenance, and Security
Study data and model drift, online performance, endpoint security, maintenance, and cost. Define monitoring signals, alarm conditions, and when to retrain or roll back.
Common mix-up: retraining automatically after a drift alarm without validation or approval, or failing to restrict who can invoke an endpoint. Treat monitoring, permissions, and maintenance as one process.
Check your MLA-C01 readiness
- How do you check that training and inference use consistent features?
- When should a drop in endpoint performance trigger a model-version rollback?
- After detecting data drift, who owns retraining, human review, and release?
Suggested review order
- Data preparation and feature quality.
- Model selection, metrics, and training refinement.
- Inference modes, deployment, and workflow orchestration.
- Monitoring, drift, security, and maintenance.
A three-step preparation plan
- Confirm the exam version. The current AWS page distinguishes MLA-C01 and MLA-C02. Confirm language, date, and code before choosing study materials.
- Use the legacy guide. If you are taking a remaining MLA-C01 language version, study the skill statements in its four domains.
- Practice and review. Group MLA-C01 practice misses by domain and cause, then revisit the relevant skills and AWS documentation.
Start practicing
If you confirmed that you are taking a remaining version of the legacy exam, use code MLA-C01 to practice by domain. For MLA-C02, use the AWS guide and exam code published for that beta.