Skip to content
Testero
Start free

Professional Machine Learning Engineer

Professional · Practice available

Find your study focus for Google’s machine learning certification with a free diagnostic and a domain-by-domain readiness report.

Start PMLE diagnostic

Official Google PMLE certification page. Facts checked .

Who this exam is for

This exam is relevant to engineers taking ML and AI solutions from an initial approach into repeatable production operation. The guide spans low-code options, shared data and model work, model scaling, serving, pipelines, and monitoring. It therefore asks for broader preparation than training an accurate model in a notebook. People working between data teams, application teams, and platform operations can use the scope to identify which parts of that lifecycle they have not yet practiced.

Google recommends: 3+ years of industry experience including 1 or more years designing and managing solutions using Google Cloud. Source.

How to prepare

  1. Organize the standard guide around a model lifecycle. Connect low-code architecture decisions to data collaboration, then follow a prototype through serving, orchestration, and monitoring. Note where business constraints would change an otherwise reasonable technical choice.
  2. Use the Testero PMLE diagnostic to identify topics for a closer look. Treat the results as a study aid, then check weak areas against the official guide instead of assuming that one diagnostic covers every objective.
  3. Practice with Google Cloud Skills Boost labs that expose deployment and operational behavior as well as model building. Record how you would investigate a pipeline failure, a serving problem, or a change in model behavior.
  4. Review the official sample questions and explain the constraints behind each answer. Return to your lifecycle map afterward and add the missing reasoning, without treating sample performance as a prediction of an exam outcome.

Exam details

There are no prerequisites. Google does not publish a PMLE passing score. Delivery is online-proctored remotely or onsite-proctored at a testing center.

Renew within the renewal eligibility period; see Google’s Renewal FAQs. The certification page does not state a validity duration.

Standard exam

Duration
Two hours
Question count
50–60
Question types
Multiple choice, Multiple select
Languages
English, Japanese
Price (USD)
$200 (plus tax where applicable)
Delivery
Online proctored or at a test center

Exam guide sections

Exam guide

Guide as of 2026-06-01

  • 1. Architecting low-code AI solutions~13%
  • 2. Collaborating within and across teams to manage data and models~16%
  • 3. Scaling prototypes into ML models~21%
  • 4. Serving and scaling models~20%
  • 5. Automating and orchestrating ML pipelines~18%
  • 6. Monitoring AI solutions~13%

Announcements and upcoming versions

  • Google reports new product branding, including the move from Vertex AI to Gemini Enterprise Agent Platform, changes to the data stack and an emphasis on Google Cloud native solutions. Use the linked current exam guide.

Official sources

Check Google before booking.

Related guides

Related certifications

All certification levels →

A diagnostic is a study signal, not an exam-outcome guarantee.