The correct option is Native integration that unifies Analytics, BigQuery, Vertex AI, and Google Ads.

This pipeline starts with Google Analytics clickstream that can be exported natively into BigQuery and joined with subscription revenue. From there Vertex AI can read data directly from BigQuery to run a generative model that produces tailored ad copy for each audience segment. Finally the results can be activated through the Google Ads API and through linked Analytics and Ads accounts. This shows an end to end flow that works smoothly because the products are designed to interoperate without custom connectors.

Open ecosystem that lets teams bring third party or open source models is not what the scenario highlights. The workflow relies on first party Google services working together rather than bringing external models into the stack.

Security by design infrastructure that protects data against breaches is always important but the scenario emphasizes seamless movement from data to modeling to campaign activation rather than security controls.

AI optimized hardware such as Cloud TPUs to reduce training cost is not indicated by the use case. The team is generating copy with Vertex AI and there is no focus on hardware selection or training cost optimization.

Full AWS Practitioner Certification Question