← AWS MLA-C01 — ML Engineer Associate

MLA-C01 Exam Overview & Strategy

MLA-C01 Exam Overview & Strategy

Certification: AWS Certified Machine Learning Engineer — Associate Code: MLA-C01 Target: ML practitioners with 1+ year hands-on SageMaker experience


Exam Format

DetailValue
Questions65 (50 scored + 15 unscored)
Duration170 minutes (~2.5 min per question)
Passing Score720 / 1000 (scaled)
Cost$150 USD
Validity3 years
FormatMultiple choice, multiple response
LanguagesEnglish, Japanese, Korean, Simplified Chinese
TestingPearson VUE (center or online proctored)

Four Domains & Weights

┌─────────────────────────────────────────────────────────────┐
│                    MLA-C01 EXAM DOMAINS                     │
├─────────────────────────────────────────────────────────────┤
│                                                             │
│  ██████████████████████████████  Domain 1: Data Prep (28%)  │
│  ████████████████████████████    Domain 2: Model Dev (26%)  │
│  ██████████████████████████      Domain 4: Monitor   (24%)  │
│  ████████████████████████        Domain 3: Deploy    (22%)  │
│                                                             │
└─────────────────────────────────────────────────────────────┘
DomainWeightFocus
1. Data Preparation for ML28%Ingestion, storage, transformation, feature engineering, data integrity
2. ML Model Development26%Algorithm selection, training, tuning, evaluation
3. Deployment & Orchestration22%Endpoints, pipelines, CI/CD, infrastructure-as-code
4. Monitoring, Maintenance & Security24%Model Monitor, drift, cost optimization, IAM, encryption

Key Difference: MLA-C01 vs MLS-C01 (Specialty)

FactorMLA-C01 (Associate)MLS-C01 (Specialty)
LevelAssociateSpecialty
FocusImplementation & operationalizationBuilding, training, tuning
Duration170 min170 min
Questions6565
EmphasisSageMaker + MLOps + BedrockDeeper ML theory + algorithms
DifficultyModerateHigher

MLA-C01 is more hands-on engineering focused — how to build, deploy, and run ML in production on AWS.


What Dominates the Exam

SageMaker is ~60%+ of the exam. Master these components:

  • SageMaker Pipelines (workflow orchestration)
  • Feature Store (online/offline feature management)
  • Model Monitor (drift detection in production)
  • Model Registry (versioning, approval workflows)
  • Endpoints (real-time, serverless, async, batch)
  • Clarify (bias detection + explainability)

Amazon Bedrock covers ~10%:

  • Foundation model access
  • Knowledge Bases (RAG)
  • Agents (multi-step automation)
  • Guardrails (content safety)

Study Strategy (6-8 Weeks)

WeekFocus
1-2Data ingestion, S3, Kinesis, Glue, EMR, feature engineering
3-4SageMaker algorithms, training, tuning, Bedrock, GenAI
5-6MLOps, Pipelines, deployment, CI/CD, Docker, monitoring
7-8Security, IAM, cost optimization, full practice exams

Allocate 50%+ of study time to hands-on SageMaker labs.


Exam Day Tips

  1. Pace yourself — ~2.5 min per question, flag and move on
  2. Eliminate first — remove obviously wrong answers
  3. Prefer AWS-native — when multiple approaches work, pick the AWS service
  4. Read scenario carefully — real-time vs batch, cost vs performance, security requirements
  5. SageMaker is usually the answer — if a question is about ML workflow, think SageMaker first

Course Alignment (Udemy — Frank Kane & Stephane Maarek)

Course SectionExam Domain
S2: Data Ingestion & StorageDomain 1 (28%)
S3: Data Transformation & Feature EngineeringDomain 1 (28%)
S4: AWS Managed AI ServicesDomain 2 (26%)
S5: SageMaker Built-In AlgorithmsDomain 2 (26%)
S6: Model Training, Tuning & EvaluationDomain 2 (26%)
S7: Generative AI FundamentalsDomain 2 (26%)
S8: Bedrock & GenAI ApplicationsDomain 2 (26%)
S9: MLOpsDomain 3 (22%)
S10: Security, Identity & ComplianceDomain 4 (24%)