4.1 Describe the Google Cloud-Recommended Steps to Successfully Implement a Transformational Gen AI Solution. GENERATIVE-AI-LEADER Practice Questions (Page 3)

Questions 11–15

1. 11expert · hard

   A gaming company wants to use gen AI to create personalized in-game narratives for players. They have a large player base with diverse preferences and need to generate unique storylines in real-time. They also have a tight budget and need to minimize infrastructure costs. Which solution is MOST suitable?

   • ADeploy a large open-source language model on dedicated GPUs for full control.
   • BUse a code generation model to write branching story scripts for each player.
   • CTrain a custom narrative generation model on player data to achieve the best personalization.
   • DUse a pre-trained text generation model with prompt engineering to create personalized narratives based on player actions.
   Reveal Solution Discussion Select an answer first
2. 12application · medium

   A logistics company wants to use gen AI to optimize delivery routes. They have real-time traffic data and need to generate optimal routes that adapt to changing conditions. They also want to provide drivers with turn-by-turn navigation. Which type of gen AI solution is MOST suitable?

   • AAn image generation model that creates visual maps of the routes.
   • BA code generation model that writes routing algorithms from scratch.
   • CA personalized user need generation solution that uses real-time data to generate adaptive route recommendations.
   • DA text generation model that produces route descriptions from traffic data.
   Reveal Solution Discussion Select an answer first
4. 13application · medium

   A customer support organization has implemented a gen AI solution that drafts responses to customer emails. They want to evaluate the solution's impact on response quality and customer satisfaction. Which approach would provide the MOST reliable evaluation?

   • ATrack the time taken to generate each response.
   • BMeasure the number of emails processed per day by the AI system.
   • CConduct A/B testing where some customers receive AI-drafted responses and others receive human-drafted responses, then compare satisfaction scores.
   • DAsk the AI system to rate its own responses for quality.
   Reveal Solution Discussion Select an answer first
5. 14application · medium

   A media company has deployed a gen AI tool that generates article headlines. They want to evaluate the tool's impact on editorial productivity and reader engagement. Which combination of metrics would provide the MOST comprehensive evaluation?

   • ATime to generate headlines and click-through rate (CTR) on articles.
   • BNumber of headlines generated and total word count.
   • CCost per API call and number of API errors.
   • DModel inference latency and GPU utilization.
   Reveal Solution Discussion Select an answer first
6. 15expert · hard

   A global bank wants to implement a gen AI solution for real-time fraud detection. They have a high volume of transactions and need sub-second latency. They also have strict data residency requirements, with data that must remain in specific countries. The bank wants to minimize operational overhead while ensuring compliance. Which deployment strategy is MOST appropriate?

   • ADeploy the model in a multi-region setup with regional endpoints to keep data within each country.
   • BDeploy the model on-premises in each country to ensure data residency, but this increases operational overhead.
   • CUse a serverless approach with a single endpoint that processes data in the cloud region closest to the user.
   • DDeploy a single global model in one region and route all traffic there.
   Reveal Solution Discussion Select an answer first