PMLEGoogle Cloud Professional Machine Learning Engineer
What the exam covers
- Architecting low-code AI solutions 46 practice questions 13%
- Collaborating to manage data and models 52 practice questions 14%
- Scaling prototypes into ML models 85 practice questions 18%
- Serving and scaling models 66 practice questions 20%
- Automating and orchestrating ML pipelines 57 practice questions 22%
- Monitoring AI solutions 44 practice questions 13%
Study guide
From Google
About PMLE
Building and running ML on Google Cloud, from BigQuery ML and AutoML through custom training on Vertex AI, pipelines, serving and monitoring. Generative AI and foundation models in Model Garden now feature alongside classic ML.
Who it's for. ML engineers who take models from notebook to production on Google Cloud.
How to prepare
- Prefer the lowest-effort option that meets the requirement: BigQuery ML or AutoML before custom training.
- Know Vertex AI Pipelines, Experiments, Feature Store and Model Registry, and what each one tracks.
- Monitoring questions separate training-serving skew from prediction drift.
Sample PMLE questions
Try these with the answers on the page, then practise the full bank.
- A regional bank wants a chat assistant that answers account and product questions. It should follow the bank's policy wording and draw on its published help articles. The team wants to configure this through a console and plain-language guidance, not by writing an intent classifier. Which Google Cloud offering is most appropriate? Architecting low-code AI solutions
- A payments company produces many builds of its fraud scoring model on Vertex AI. Auditors want a record of every build, the one currently serving live traffic, and the evaluation numbers for each build. Which capability should the team use? Collaborating to manage data and models
- A lender must tell each rejected applicant the main factors behind that specific decision. The model is a gradient boosted classifier deployed on Vertex AI. Which technique provides this? Scaling prototypes into ML models
- A data science lead notices that her classifier scores 0.98 accuracy on the examples used to fit it but only 0.71 on a holdout set drawn from the same population. Which diagnosis fits best? Scaling prototypes into ML models
- An engineer fine-tuning a transformer language model on a single accelerator keeps hitting out-of-memory errors, and the batch size they need is out of reach. Which two changes most directly lower peak memory while keeping training quality? Scaling prototypes into ML models
- A fintech app calls a fraud scorer for every card swipe and must answer within 100 milliseconds at the 95th percentile, with traffic that spikes during evenings. Which hosting choice best fits? Serving and scaling models
- While training a churn predictor, the loss on the training split keeps falling, but the loss on a held-out validation split has risen over the last several evaluation rounds. What is the most appropriate stopping rule? Scaling prototypes into ML models
- In a housing price model, floor area, room count and built-up area are almost perfectly correlated, and the coefficient estimates swing wildly between retrains. Which remedy addresses this redundancy? Scaling prototypes into ML models
- All 40 sample questions
Questions people ask
Is this PMLE practice exam free?
Yes. All 350 questions, the timed mock exam and the explanations are free, and there's no account to create.
Are these real exam questions?
No. Real exam questions are confidential under the Google Cloud candidate agreement. These are practice questions written to cover the same domains in the same scenario style, each with an explanation of why the answer is right.
How is the PMLE mock exam scored?
Questions are drawn from every domain in proportion to the exam guide, and you get a raw percentage with a breakdown by domain. Google Cloud uses scaled scoring, so treat 75% or more here as a sign you're ready rather than a predicted score.
Where is my progress saved?
In your browser, on this device. Nothing is sent to us apart from the answers you check, and clearing site data resets it.