Directory / Google Cloud

Professional Machine Learning Engineer

A senior ML engineering and MLOps exam on Google Cloud, with generative AI additions such as model selection, tuning and LLM evaluation. It does not test agent building.

What it signals

Practical assessment: Scenario-based multiple choice. Proctored scenario-based multiple choice/multiple select with code snippets to read, no interactive items.
Identity assurance: Proctored with identity verification. Proctored: online-proctored (Pearson VUE OnVUE) or onsite at a Pearson VUE test centre. Registration via CertMetrics (cp.certmetrics.com/google). Third-party source says delivery moved from Kryterion to Pearson VUE on 2026-03-02.
Depth: Professional. Professional level.
Experience expected: 2 to 3 years. Google recommends 3+ years industry experience including 1+ on Google Cloud.
Portability: Mixed. ML/MLOps concepts transfer but most objectives are framed around specific Google Cloud products.
Verifiability: Public badge or registry. Public Credly badge and searchable directory.

For hiring managers

Exam was revised effective 2026-06-01; product names changed from Vertex AI to Gemini Enterprise Agent Platform. Holders certified before that date sat the older Vertex AI-named version.
Despite third-party claims that the new version is agent-heavy, the official June 2026 exam guide contains NO objectives about building agents; 'Agent Platform' appears only as the product name. It remains an ML engineering/MLOps exam with gen AI additions (model selection in Model Garden, Gemini tuning, LLM-as-a-judge evaluation, Model Armor).
Multiple-choice only; does not prove coding ability.
Section names and weights are unchanged in structure from the prior version (6 sections).

Verifying it

Digital badge issued via Credly (Open Badges); employer opens the candidate's live Credly badge URL (shows name, credential, issue and expiry dates). Opt-in searchable Google Cloud Skills Directory on Credly: https://www.credly.com/organizations/google-cloud/directory

What it covers

Published exam outline

Architecting low-code AI solutions 13%
Collaborating within and across teams to manage data and models 16%
Scaling prototypes into ML models 21%
Serving and scaling models 20%
Automating and orchestrating ML pipelines 18%
Monitoring AI solutions 13%

By topic

GenAI apps: Some (10 to 20%). Objective 1.2, fine-tuning foundation models, and gen AI evaluation/monitoring bullets add up to roughly 10-20%.
Agents: Not covered. No agent-building, tool-use or orchestration objectives appear in the official guide.
Classical ML: Core focus (over 35%). Sections 1.1, 2.2-2.3 and all of section 3 (21%) concern building, training, tuning and evaluating models, together over 35%.
Operations: Core focus (over 35%). Sections 4, 5 and 6 (serving, pipelines, monitoring) total about 51%.
Data: Some (10 to 20%). Objective 2.1 and training data organisation/ingestion in 3.2 are roughly 10-15%.
Responsible AI: Touched on (under 10%). Objective 6.1 (security, bias, explainability) and PII handling are under 10%.

Exam details

Who the provider says it is for. ML Engineers who build, evaluate, productionize and optimize AI solutions on Google Cloud, including solutions based on foundational models; strong programming skills and experience with data platforms.

Format: multiple choice; multiple select
Delivery: Proctored: online-proctored (Pearson VUE OnVUE) or onsite at a Pearson VUE test centre. Registration via CertMetrics (cp.certmetrics.com/google). Third-party source says delivery moved from Kryterion to Pearson VUE on 2026-03-02.
Code or hands-on work: No code writing, no labs. Official note: 'The exam does not directly assess coding skills. If you have a minimum proficiency in Python and SQL, you should be able to interpret any questions with code snippets.' Candidates must READ code snippets.
Prerequisites: None
Recommended experience: 3+ years of industry experience including 1+ years designing and managing solutions using Google Cloud
Renewal: Renew within renewal eligibility period by retaking the exam. No shorter renewal exam or continuing-education route was listed for PMLE in the Renewal FAQs content retrieved (those exist for Cloud Architect, Data Engineer, ACE, CDL).

Preparing

Not confirmed

These details were not available from an official source at the time of research.

passingScore (not published)
validity period (2 years) not confirmed on an official page in this session
exact renewal window for PMLE
exact go-live date of new exam version beyond the guide's 'as of June 1, 2026' title
examCode

Sources (5)