You work for a retail company that collects customer data from various sources:Online transactions: Stored...

Comprehensive and Detailed In-Depth Explanation:

The pipeline must extract diverse data types and load them into systems optimized for analysis and ML. Let’s assess:

Option A: Cloud SQL for transactions keeps data relational but isn’t ideal for analysis/ML (less scalable than BigQuery). BigQuery for feedback is fine but skips staging. Cloud Storage for streaming social media loses real-time context and requires extra steps for analysis.

Option B: BigQuery for transactions (via export from MySQL) supports analysis/ML with SQL. Cloud Storage stages feedback text files for preprocessing, then BigQuery ingestion. Pub/Sub and Dataflow stream social media into BigQuery, enabling real-time analysis—optimal for all sources.

Option C: Cloud Storage for all data is a staging step, not a final solution for analysis/ML, requiring additional pipelines.