AWS Machine Learning Certification: What Replaced the Specialty Exam
If you searched for the AWS Certified Machine Learning Specialty exam this year, you found a retired credential. AWS closed registration for MLS-C01 on March 31, 2026, and pointed every new candidate toward a different exam with a different focus. If you are an experienced IT professional weighing this certification for a promotion or a move into ML operations, here is what exists now and how to plan around it.
Ultimate IT Courses runs instructor-led AWS certification training built around current exam objectives. Book a training consultation to map out which AWS Machine Learning exam version fits your timeline.
The Specialty Exam Is Gone
MLS-C01 tested candidates on building and tuning machine learning models from scratch, a skill set built around research-style data science work. AWS retired it because most ML work on its platform no longer looks like this. Teams build on managed services and foundation models rather than training custom models line by line. AWS shifted its credential to match, and if you hold the old Specialty certification, it stays valid for three years from the date you earned it.
What Replaced It
The successor is the AWS Certified Machine Learning Engineer, Associate level, exam code MLA-C01. It validates your ability to build, deploy, and maintain ML solutions and pipelines on AWS, with a heavier weight on production skills than the old exam carried. The change reflects a broader industry pattern, where employers want engineers who ship and monitor ML systems, not only people who design an algorithm on a whiteboard.
Here is where the timeline gets tight. MLA-C01 itself is on its way out. AWS opened beta registration for its replacement, MLA-C02, on September 1, 2026, and set September 28, 2026, as the last day to sit MLA-C01 in English. The standard version of MLA-C02 arrives in early 2027. If you plan to test soon, confirm which version you are registering for before you book a seat.
Exam Domains and What They Cover
According to the official AWS exam guide, MLA-C01 runs 65 questions over 170 minutes, with a passing score of 720 out of 1,000, and a fee of 150 USD. The exam splits across four domains: data preparation for machine learning at 28 percent, ML model development at 26 percent, deployment and orchestration of ML workflows at 22 percent, and ML solution monitoring, maintenance, and security at 24 percent. In practice, this means you need working knowledge of ingesting and validating data, training and tuning models in Amazon SageMaker, choosing deployment infrastructure and configuring auto scaling, and setting up CI/CD pipelines for ML workflows. MLA-C02 keeps this same four-domain structure but updates the skills within each domain to cover foundation models, large language models, and agentic AI workflows alongside traditional ML engineering.
Who Should Pursue This Certification
AWS recommends at least one year of experience in a related role, such as backend development, DevOps, data engineering, or data science, plus hands-on time with SageMaker and other AWS ML services. This certification fits three groups well:
- Backend or DevOps engineers moving toward ML infrastructure work
- Data engineers who already build pipelines and want AWS-specific credentials
- Data scientists shifting from model-building toward production deployment and monitoring
If none of these describe your background yet, an associate-level AWS credential like the Solutions Architect Associate gives you a stronger foundation first.
How to Plan Your Study Timeline
Given the transition, decide early which version to target. If your schedule allows you to prepare and test before September 28, 2026, MLA-C01 remains available in English, and study materials for it are mature. If your timeline runs longer, plan for MLA-C02 once it leaves beta, since AWS will not keep both versions active. Either path tests the same core domains, so the study foundation carries over regardless of which exam code you end up sitting.
Where This Fits in Your Career
Job Bank Canada tracks machine learning engineers under the broader data scientist occupation code, and it currently rates the outlook as moderate in provinces like British Columbia and Alberta and more limited in Ontario and the Toronto region. Demand ties closely to how fast organizations adopt AI, cloud infrastructure, and automation, so the certification carries more weight when paired with real production experience than as a standalone credential.
For an experienced IT professional already working in AWS environments, this certification signals you operate ML systems at production scale instead of only prototyping them.
Where Ultimate IT Courses Fits In
Ultimate IT Courses runs instructor-led AWS certification programs with small class sizes, so you work through SageMaker, deployment pipelines, and monitoring tools with direct instructor access rather than a static video course. Book a training consultation to map out which AWS Machine Learning exam version fits your timeline and your current experience.
