Professional Data Engineer Practice Exam: Test Your Knowledge 2025

Your company is migrating a data warehouse from on-premises to Google Cloud. The warehouse contains 50 TB of historical data that needs to be queried frequently using SQL. The data is updated daily with batch loads. Which Google Cloud service is most appropriate for this use case?

You are designing a real-time data pipeline to process IoT sensor data from thousands of devices. The data must be ingested, processed, and made available for analysis with sub-second latency. Which combination of services should you use?

A financial services company needs to store transaction data with strong consistency, support ACID transactions across multiple rows and tables, and scale globally. The database must support SQL queries. Which service should they use?

Your team needs to design a data lake architecture on Google Cloud that can store structured, semi-structured, and unstructured data. The solution must support both batch and streaming ingestion, be cost-effective for long-term storage, and allow data to be processed by multiple analytics tools. What architecture should you implement?

You need to load 10 GB of CSV files from Cloud Storage into BigQuery daily. The data has inconsistent formats and requires validation and transformation before loading. What is the most efficient approach?

Your Dataflow pipeline is processing streaming data but experiencing high latency during peak hours. You've noticed that some workers are overloaded while others are idle. What should you do to optimize the pipeline?

You have a Cloud Composer (Airflow) environment running multiple DAGs that orchestrate BigQuery and Dataflow jobs. Some DAGs are failing intermittently due to transient network errors. How should you make the workflows more resilient?

Your company needs to process Apache Spark jobs on Google Cloud. The jobs run for 2-3 hours daily and require specific library dependencies. You want to minimize operational overhead and cost. What solution should you implement?

You are implementing a CI/CD pipeline for your data processing workflows that include BigQuery stored procedures, Dataflow templates, and SQL transformations. What is the best approach to version control and deployment?

Your organization has deployed a machine learning model using Vertex AI for predictions. You need to monitor the model for prediction drift and data quality issues. What should you implement?

You have trained a TensorFlow model for image classification and need to deploy it for real-time predictions with low latency (under 100ms) and the ability to scale to thousands of requests per second. Which deployment option should you choose?

Your data science team has developed multiple ML models using different frameworks (TensorFlow, PyTorch, scikit-learn). You need to create a unified MLOps pipeline for training, versioning, and deployment. What approach should you take?

You need to perform batch predictions on 500 GB of data stored in BigQuery using a trained model. The predictions are not time-sensitive and should be cost-effective. What is the best approach?

Your ML model training job in Vertex AI is taking too long to complete. You're training a deep learning model with large image datasets stored in Cloud Storage. What optimization strategies should you implement?

You need to implement A/B testing for two versions of a deployed ML model to compare their performance before fully rolling out the new version. Which Vertex AI feature should you use?

You are responsible for ensuring data quality in a BigQuery data warehouse. Users report that some dashboards show incorrect aggregations. What approach should you implement to prevent data quality issues?

Your BigQuery queries are running slower than expected and incurring high costs. After investigation, you find that many queries scan entire tables even when filtering on specific dates. What optimization should you implement?

Your data pipeline processes sensitive customer PII. You need to implement security controls to ensure data is encrypted, access is audited, and sensitive fields are protected. What combination of security measures should you implement?

You need to set up monitoring and alerting for your data pipeline to detect failures and performance degradation. The pipeline uses Dataflow, BigQuery, and Cloud Storage. What monitoring strategy should you implement?

Your organization needs to implement disaster recovery for critical BigQuery datasets and ensure business continuity with an RTO of 4 hours and RPO of 1 hour. What approach should you take?