Assembling the services a simple transformation pipeline needs
For basic transformation pipelines on Google Cloud, the Associate Data Practitioner exam expects you to match the transformation need (batch vs streaming, code vs no-code, SQL vs programmatic) to the right service. Common building blocks include Cloud Data Fusion, Dataflow, Dataproc, and BigQuery itself, often orchestrated with Cloud Composer or Workflows.
1 · Learn the must-know
- Cloud Data Fusion is the go-to choice for a visual, no-code/low-code drag-and-drop ETL/ELT pipeline builder using pre-built connectors and transformations.
- Dataflow is the preferred fully managed service for both batch and streaming data transformations using Apache Beam, ideal when you need custom code-based, scalable pipelines with autoscaling.
- Dataproc is best when you have existing Hadoop/Spark workloads or need Spark-based transformations and want more control over cluster management than Dataflow provides.
- BigQuery supports in-place ELT transformations via SQL (including scheduled queries and BigQuery Data Transfer Service), which is the simplest option when data is already loaded and transformations can be expressed in SQL.
- Cloud Composer (managed Apache Airflow) is used to orchestrate multi-step pipelines across multiple GCP services, not to perform the transformation itself.
- Cloud Workflows is a lighter-weight orchestration option for simpler, serverless service-to-service orchestration compared to Composer's full Airflow-based DAG orchestration.
2 · Check your understanding
A data practitioner at a retail company needs to build a pipeline that reads CSV files landing in Cloud Storage, renames a few columns, filters out incomplete rows, and writes the results to BigQuery. The practitioner wants to design the transformation logic visually, without writing custom pipeline code, and prefers a managed service with prebuilt connectors. Which product should the practitioner use to implement this pipeline?
What you have tried across GCP ADP's objectives, not a readiness score.
Data Preparation and Ingestion~30% of the exam0 of 8 tried
Data Analysis and Presentation~27% of the exam0 of 12 tried
Data Pipeline Orchestration~18% of the exam0 of 9 tried
Data Management~25% of the exam0 of 12 tried
3 · Keep going
Ready for more? Take a weighted mock or try free practice questions.