You need to orchestrate a pipeline with several Google Cloud services: a batch Dataflow job,...

When the requirement specifies a "lightweight" and "serverless" orchestration for Google Cloud APIs with "minimal operational overhead," Cloud Workflows is the preferred choice over Cloud Composer.

Lightweight and Serverless: Cloud Workflows is a fully managed, HTTP-based orchestration service that scales to zero and has no base cost. It is designed specifically to chain Google Cloud services together using YAML or JSON.

Operational Overhead: Unlike Cloud Composer (which requires managing a Kubernetes-based environment and has a minimum running cost), Workflows is truly serverless with no infrastructure to manage.

Service Integration: Workflows has built-in connectors for Dataflow, BigQuery, and Vertex AI, making it ideal for simple sequential logic.

Correcting other options:

B (Cloud Composer): While it can handle this logic, it is not "lightweight." It is better suited for complex data engineering pipelines with non-GCP dependencies.

C (Compute Engine): This is not serverless and requires significant operational overhead to manage the VM and cron state.

D (Dataproc/Oozie): This is a legacy Hadoop-based orchestration tool and is definitely not lightweight or serverless.