Exam Overview
What the Google Cloud Professional Data Engineer exam actually tests, a week-by-week plan for the seven domains, and the key services to know before you sit it.
What the exam tests
The Google Cloud Professional Data Engineer certification validates your ability to design, build, secure, and optimize data processing systems on GCP. The exam is 2 hours, ~50-60 questions, and tests practical architecture decisions rather than trivia recall.
Questions typically present a scenario with constraints (cost, latency, compliance) and ask you to pick the most appropriate architecture. Focus on understanding why one option is better given specific tradeoffs.
Key Services at a Glance
Click any service to expand a study card — overview, key features, when to use, and exam tips.
Suggested Study Plan
Show the 7-week plan
Week 1 — Data Ingestion
Focus on Pub/Sub messaging guarantees, Dataflow's unified batch/stream model, Datastream CDC, Storage Transfer Service, and patterns for choosing between ingestion tools based on latency and cost requirements.
Week 2 — Data Processing
Deep-dive into Apache Beam concepts (windowing, triggers, watermarks), Managed Service for Apache Spark cluster tuning, ephemeral vs. persistent clusters, and Managed Service for Apache Airflow DAG design patterns including ExternalTaskSensor.
Week 3 — Data Storage
Cover GCS lifecycle policies, Bigtable row key design and hotspot avoidance, Spanner global transactions and interleaving, Cloud SQL HA and PITR, and Firestore Native vs. Datastore mode.
Week 4 — Analytics
Master BigQuery partitioning and clustering, materialized views, INFORMATION_SCHEMA queries for cost analysis, on-demand vs. flat-rate pricing, BI Engine, and Looker vs. Looker Studio selection.
Week 5 — Operations
Study Dataflow Flex Templates and operational metrics (system lag, data freshness), Managed Service for Apache Airflow environment tuning, cost attribution with labels, and Cloud Logging log sinks and audit logs.
Week 6 — Machine Learning
Cover Gemini platform end-to-end (training, tuning, deployment), Feature Store freshness, BigQuery ML (ARIMA_PLUS, AutoML), model monitoring and drift detection, Document AI, and distributed training patterns.
Week 7 — Security, Compliance & Review
Study IAM best practices, VPC Service Controls perimeter design, Sensitive Data Protection tokenization and de-identification, CMEK, Knowledge Catalog data lineage, and GDPR crypto-shredding. Spend the final 2-3 days on the 4 full-length mock exams — timed 2-hour, 50-question simulations scored pass/fail with a per-domain weak-area report.
Where to go next
The seven study domains go deeper than this page, and each guide ends with three practice questions. Once the domains feel solid, test yourself under exam conditions with a full-length mock exam.