Which GCP service is best for real-time analytics on streaming data?

Correct answer: Dataflow

Option C is correct because Google Cloud Dataflow is a fully managed stream and batch processing service built on Apache Beam, designed specifically for real-time data pipeline processing and analytics on streaming data with low latency and autoscaling. Option A is incorrect because BigQuery is a serverless data warehouse optimized for interactive SQL analytics on stored datasets, not for processing live streaming pipelines directly, though it can receive streamed inserts as a destination. Option B is incorrect because Dataproc is a managed Apache Hadoop and Spark service suited for large-scale batch processing workloads, not purpose-built for real-time streaming analytics. Option D is incorrect because Pub/Sub is a messaging and event ingestion service that acts as the pipeline for delivering streaming messages, but it does not perform analytics or data transformations itself.