AWS Certified Machine Learning – Specialty

MLS-C01 · ● RETIRING (Last day: March 31, 2026) · Expert · Amazon Web Services

RETIREMENT ALERT: This certification ends March 31, 2026. AWS recommends candidates transition to:

• AIF-C01 (AI Practitioner – Foundational) for entry-level AI/ML knowledge
• MLA-C01 (Machine Learning Engineer – Associate) for hands-on ML implementation
• AIG-C02 (Generative AI Developer – Professional, emerging 2025+) for LLM applications

New registrations are no longer accepted. Existing certifications remain valid for 3 years from issuance date.

Exam facts

Vendor source — AWS Certified Machine Learning – Specialty ↗

Official exam guide — AWS Certified Machine Learning – Specialty Exam Guide ↗

Exam objectives — MLS-C01 Exam Blueprint ↗

About

The AWS Certified Machine Learning – Specialty (MLS-C01) validates expertise in building, training, tuning, and deploying machine learning models on AWS. Designed for ML engineers, data scientists, and ML architects with 1–2+ years of hands-on experience, this exam covers the full ML lifecycle: data engineering, exploratory data analysis, model selection, hyperparameter tuning, and production deployment via AWS SageMaker. Originally launched around 2016, the exam is being sunset March 31, 2026, as AWS shifts toward the newer MLA-C01 (ML Engineer Associate) for technical practitioners and AIF-C01 (AI Practitioner Foundational) for broader AI/ML awareness. Existing MLS-C01 certifications retain validity for 3 years post-issuance.

Domain context — Data / AI

ML specialty expertise within the AWS ecosystem, spanning data pipelines, feature engineering, model training, ensemble methods, and MLOps at scale.

Read full deep dive — Data / AI Domain →

Read full deep dive — AWS Ecosystem →

Topics covered

The MLS-C01 exam blueprint covers four weighted domains:

• Data Engineering (20%) — Data sources, ingestion, storage, pipeline design using AWS Glue, Kinesis, EMR, Athena, Redshift; data quality and governance
• Exploratory Data Analysis (24%) — Statistical analysis, feature engineering, data visualization, handling missing/imbalanced data, using SageMaker Data Wrangler and Feature Store
• Modeling (36%) — Algorithm selection, hyperparameter tuning via SageMaker Hyperparameter Tuning, ensemble methods (XGBoost, Factorization Machines, Linear Learner), deep learning on SageMaker (TensorFlow, PyTorch, MXNet), model evaluation, cross-validation, bias/fairness checks with SageMaker Clarify
• ML Implementation and Operations (20%) — SageMaker training/inference job orchestration, multi-model endpoints, batch transform, async inference, model hosting, A/B testing, monitoring via SageMaker Model Monitor, debugging with SageMaker Debugger, cost optimization, CI/CD for ML

Source: AWS MLS-C01 Exam Blueprint ↗

Common skills at Data / AI · Expert

Shared expertise for the Data / AI domain at Expert / Specialist level — not specific to this cert.

• Full ML lifecycle ownership (data ingestion through model deprecation)
• Advanced feature engineering and dataset optimization at scale
• Hyperparameter tuning and model selection across classical and deep learning frameworks
• Ensemble methods (stacking, boosting, bagging) and advanced model evaluation
• Deep learning on GPU/TPU with TensorFlow, PyTorch, MXNet
• MLOps discipline: experiment tracking, model registry, automated retraining pipelines
• Model fairness, bias detection, and explainability (SHAP, LIME)
• Production inference optimization: latency, throughput, cost
• Monitoring, alerting, and automated remediation in production ML systems

Recommended courses at Data / AI · Expert

Course-selection rule: The Udemy course by Kane & Maarek is the most popular and frequently updated for MLS-C01; AWS Skill Builder is the official path but less detailed than third-party courses. Given the retirement date (March 31, 2026), all of these courses should be treated as legacy after that date—consider MLA-C01 courses instead for new learners.

Practice exams

Note: With the March 31, 2026 retirement approaching, practice exam availability may decline. Verify current pricing and access before purchase.

Books

Book rule: The Sybex guide (Subramanian & Natu, 2022) is the most authoritative; the authors hold principal ML roles at AWS. Post-retirement (after March 31, 2026), these books become reference materials only, not primary exam prep.

Typical job titles at Data / AI · Expert

Senior ML Engineer · ML Specialist · AI/ML Solutions Architect · ML Tech Lead · Principal Machine Learning Engineer · ML Research Engineer

(Job titles drawn from current job-board postings requiring or preferring this cert.)

Salary

Salary rule: MLS-C01 is a specialty cert; salary data reflects senior ML engineer roles. Most certified professionals hold "ML Engineer," "ML Architect," or "Data Science Lead" titles. Regional data varies; US San Francisco/NYC command +20–40% premium.

Skills validated

Cert-specific technologies and frameworks tested by MLS-C01.

• Amazon SageMaker — Studio, Notebooks, Training jobs, Hyperparameter Tuning, Multi-model endpoints, Async inference, Batch Transform, AutoML, JumpStart, Model Monitor, Clarify, Debugger, Ground Truth, Model Registry, Pipelines
• AWS data services — Glue (ETL), Kinesis (streaming), Athena (SQL analytics), EMR (Spark), Redshift (data warehouse), S3, DynamoDB
• ML algorithms & frameworks — XGBoost, Factorization Machines, Linear Learner, Seq2Seq, Object Detection, Image Classification, K-Means, Principal Component Analysis (PCA)
• Deep learning frameworks — TensorFlow, PyTorch, MXNet (on SageMaker)
• Feature engineering & data preparation — SageMaker Data Wrangler, Feature Store, Data Pipeline construction, missing data handling, class imbalance strategies
• Model evaluation & validation — Cross-validation, A/B testing, confusion matrix, precision/recall, ROC-AUC, evaluation metrics per problem type
• Bias & fairness — SageMaker Clarify, bias detection, fairness metrics, model explainability (SHAP, LIME)
• MLOps & production — Model hosting, inference optimization, cost optimization, monitoring, logging, CI/CD integration
• AWS security & compliance — VPC/security groups, IAM roles, KMS encryption, data privacy, HIPAA/PCI-DSS considerations

Related certifications

• Successor (recommended for new learners): AWS Certified Machine Learning Engineer – Associate (MLA-C01) ↗ — Lower level, more current, not retiring
• Complementary foundational: AWS Certified AI Practitioner (AIF-C01) ↗ — Entry-level AI/ML awareness
• Emerging successor (production AI): AWS Certified Generative AI Developer – Professional (AIG-C02, TBD) — LLM deployment, agents, RAG
• Stacks with: AWS Certified Solutions Architect – Professional (SAP-C02) ↗ — Complementary for full-stack AWS expertise
• Predecessor (now retired): AWS Certified Machine Learning – Specialty (MLS-C01 older exams, e.g., MLS-C00) — All legacy versions consolidated into current MLS-C01, which itself is retiring
• Vendor overview: AWS Vendor Overview ↗

Migration path (for current MLS-C01 holders)

If you hold MLS-C01 (or are considering whether to pursue it before March 31, 2026):

1. Immediate (before March 31, 2026): Complete MLS-C01 if you've already started study. The cert is still valid and worth obtaining.
2. Post-retirement transition:
   • If your role is MLOps-focused → pursue MLA-C01 (ML Engineer Associate) for AWS-native ML implementation
   • If your role is AI application-focused (LLMs, RAG, agents) → watch for AIG-C02 (Generative AI Developer – Professional) launch (2025–2026)
   • If your role is broad cloud architecture → combine MLA-C01 with SAP-C02 (Solutions Architect – Professional)
   • If you're starting from scratch → skip MLS-C01 entirely; start with AIF-C01 (foundational) → MLA-C01 (associate)

Why the change? AWS consolidated ML expertise across three pathways: foundational (AIF-C01), practitioner (MLA-C01), and emerging generative AI (AIG-C02). The old MLS-C01 tried to be both specialist and expert; the new paths are role-aligned.

Sources

• AWS Certified Machine Learning – Specialty ↗
• AWS MLS-C01 Exam Blueprint PDF ↗
• AWS Certification Exam Guides ↗
• AWS Skill Builder MLS-C01 Exam Prep Plan ↗
• Udemy: Frank Kane & Stephane Maarek AWS ML Specialty ↗
• A Cloud Guru AWS ML Specialty ↗
• Pluralsight AWS ML Specialty Path ↗
• Codecademy MLS-C01 Path ↗
• Tutorials Dojo MLS-C01 Practice Exams ↗
• Whizlabs AWS ML Specialty Practice Tests ↗
• Sybex: AWS Certified Machine Learning Study Guide (Subramanian & Natu) ↗
• Glassdoor: Senior Machine Learning Engineer Salary 2026 ↗
• Robert Half 2026 Salary Guide – Technology ↗
• Levels.fyi: Senior Machine Learning Engineer Salaries ↗
• AWS MLS-C01 Retirement Announcement ↗
• Data / AI Domain Deep Dive ↗

Last verified: 2026-05-01
Parent ecosystem: AWS Ecosystem →
Parent domain: Data / AI Domain →
Vendor overview: AWS Overview →

Status summary

RETIRING CERT — Last day to test: March 31, 2026. This document is current as of May 1, 2026, but the exam is no longer available for new registrations. Existing MLS-C01 certifications issued before the retirement date remain valid for 3 years. Learners should prioritize MLA-C01 (ML Engineer Associate) for current AWS ML engineering certification, or AIF-C01 (AI Practitioner) for foundational AI/ML knowledge. This cert remains valuable for career progression and skill validation if obtained before the deadline.