Chapter 2 of 6 · Google Cloud Professional Machine Learning Engineer
Collaborating within and across teams to manage data and models
Preprocessing, Feature Store, notebooks and experiment tracking.
- 4topics
- 20 minfast track
- 16%of the exam
Topics in this chapter
Take them in order, or jump to the one you need.
- 1Organizing, Exploring and Preprocessing ML DataOrganize tabular, text and image data for ML and pick BigQuery, Dataflow, Apache Spark or in-memory Python for preprocessing by…5 min
- 2Feature Store and Protecting Sensitive DataCreate and share features in Gemini Enterprise Agent Platform Feature Store on BigQuery, and protect PII with de-identification…5 min
- 3Prototyping in Workbench and Colab EnterpriseChoose Agent Platform Workbench or Colab Enterprise, secure and share notebooks, build PyTorch, scikit-learn or JAX models and…5 min
- 4Tracking Experiments, Evaluating Models and LineageTrack runs with Experiments on Agent Platform, evaluate predictive and gen AI models with metrics and LLM-as-a-judge, and record…5 min
About this chapter
What this chapter covers
Collaborating within and across teams to manage data and models is about 16% of the Google Cloud Professional Machine Learning Engineer exam. It covers the shared groundwork of ML on Gemini Enterprise Agent Platform (the platform formerly called Vertex AI): preparing data so several teams can use it, prototyping in secure notebooks, and recording experiments so results can be compared and reproduced.
It covers exploring and preprocessing data (organizing tabular, text and image data; choosing BigQuery, Dataflow, Apache Spark or in-memory Python by scale; building shared features in Gemini Enterprise Agent Platform Feature Store; protecting PII), prototyping in notebooks (Agent Platform Workbench and Colab Enterprise, collaboration and security, PyTorch, scikit-learn and JAX, Model Garden models) and tracking experiments (Experiments, Pipelines and Kubeflow Pipelines; evaluating predictive and gen AI models, including LLM-as-a-judge; artifacts, versions and lineage with ML Metadata).
The exam has 50 to 60 multiple choice and multiple select questions in 2 hours. Questions here usually describe a team, a data size and a constraint such as privacy or reproducibility, then ask "What should you do?".
How the topics fit together
Data work sets what every model can learn, notebooks turn ideas into prototypes, and experiment tracking makes the winning prototype reproducible. Privacy controls apply at every step.
Pick the tool by data size and team skill. Data in BigQuery and a SQL team points to BigQuery; very large or streaming data points to Dataflow; existing Spark code points to Apache Spark; data that fits in memory points to pandas in a notebook.
How to use these pages
Each topic page has the same shape, so you can skim for what you need.
- What it is: the idea in plain words.
- How it works: the mechanism, with tables and code examples.
- What to remember for the exam: the facts to memorise and how questions are phrased.
- Common mistakes: the classic traps.
- Check yourself: three questions to test your understanding.
Ready for more practice?
The full Google Cloud Professional Machine Learning Engineer question bank has 749 questions, with timed practice exams and an explanation for every answer.
This study guide is independent and is not affiliated with, endorsed by, or sponsored by Google Cloud. Vendor names, certification names and exam codes are trademarks of their respective owners and are used for identification only.