What is AWS Certified Machine Learning Engineer — Associate (MLA-C01)?
MLA-C01 is an associate-level AWS certification focused on building, deploying, and operating ML solutions and pipelines on AWS, with a strong emphasis on Amazon SageMaker and practical MLOps.
If you want the fastest “what should I learn?” view, start with the Syllabus.
What score do you need to pass MLA-C01?
AWS uses a scaled score (100–1000). The minimum passing score is 720.
How many questions and how much time?
- 65 questions
- 130 minutes
- Multiple-choice and multiple-response
Do you need to code for MLA-C01?
You don’t need to write production code during the exam, but you should be comfortable with:
- The ML lifecycle (data prep → training → evaluation → deployment → monitoring)
- Common MLOps concepts (versioning, CI/CD, monitoring, retraining triggers)
- Choosing the right AWS services and endpoint types for a scenario
What AWS services should you know for MLA-C01?
At a high level, expect to see:
- Amazon SageMaker (training, endpoints, pipelines, model registry, monitoring)
- Data prep and ETL tools (for example, AWS Glue, SageMaker Data Wrangler)
- Storage and data sources (Amazon S3, plus common data stores)
- Observability and governance (CloudWatch, CloudTrail, cost tooling)
- Security primitives (IAM, encryption, VPC basics)
Use the Cheatsheet for a service-by-use-case map.
How long should you study for MLA-C01?
Typical ranges (varies with hands-on experience):
- Strong SageMaker + ML background: 40–60 hours
- Some AWS and some ML, but not both deeply: 60–90 hours
- New to ML engineering on AWS: 90–120+ hours
Pick a schedule you can sustain: 30/60/90-day Study Plan →.
Is MLA-C01 closer to “data science” or “engineering”?
More engineering. The emphasis is on operationalizing ML: data pipelines, training and tuning workflows, deployment endpoints, CI/CD, monitoring, cost management, and security.
How do you practice effectively for MLA-C01?
Follow a loop:
- Read one task in the Syllabus
- Drill that task in Practice
- Write 3–5 “miss rules” from what you got wrong
- Re-drill weak tasks 48–72 hours later (spaced repetition)