Which Google Cloud service is specifically designed for building and deploying machine learning models?

Correct answer: Vertex AI

Option D, Vertex AI, is correct because it is Google Cloud's unified platform purpose-built for the entire machine learning lifecycle, including dataset management, model training, hyperparameter tuning, model evaluation, and deployment to serving endpoints. Option A, Cloud Composer, is a managed Apache Airflow service used for orchestrating data pipelines and workflows, not for building or deploying ML models directly. Option B, Cloud Dataflow, is a managed Apache Beam service for stream and batch data processing, typically used for ETL and data preparation rather than model training or serving. Option C, Cloud Dataproc, is a managed Spark and Hadoop service optimized for large-scale data processing and analytics workloads, not specifically designed as an end-to-end ML platform.