2.4 Describe How Google Cloud Empowers Developers to Build with AI. GENERATIVE-AI-LEADER Practice Questions (Page 5)

Questions 21–25

1. 21application · medium

   A healthcare organization wants to build an internal assistant that answers clinical policy questions by retrieving from their policy documents and generating responses that include source references. They want to minimize custom development. What should they use?

   • AModel Garden to fine-tune a model on policy documents
   • BAgent Search with prebuilt RAG
   • CAgent Platform AutoML to train a text classifier
   • DRAG APIs to build a custom pipeline from scratch
   Reveal Solution Discussion Select an answer first
2. 22expert · hard

   A large enterprise wants to build an AI assistant that can answer HR policy questions and also initiate leave requests through their HR system. They have a team of developers but want to minimize custom ML work. They also need the assistant to provide answers grounded in the official policy documents. Which approach best meets these requirements?

   • AUse Agent Search alone to provide policy answers and leave request links
   • BBuild a custom agent on Agent Platform, using Agent Search with prebuilt RAG for policy answers and a tool to call the HR system
   • CUse Agent Platform AutoML to train a model that both answers and acts
   • DUse RAG APIs to build a custom RAG pipeline and integrate with the HR system
   Reveal Solution Discussion Select an answer first
4. 23application · medium

   A law firm wants to let its attorneys ask natural-language questions across thousands of legal documents and receive answers with references to the source documents. The firm has no ML team. What should they use?

   • AAgent Search with prebuilt RAG
   • BModel Garden to fine-tune a model on legal documents
   • CAgent Platform AutoML to build a document classifier
   • DRAG APIs to build a custom retrieval system from scratch
   Reveal Solution Discussion Select an answer first
5. 24foundation · easy

   Which scenario is a typical use case for Agent Search?

   • AAutomating the deployment of applications to a Kubernetes cluster
   • BTraining a custom image classification model on a labeled dataset
   • CMonitoring cloud infrastructure for performance issues
   • DEnabling employees to search across internal documents to find answers quickly
   Reveal Solution Discussion Select an answer first
6. 25foundation · easy

   Which capability is a key benefit of using Model Garden for developers?

   • AThe ability to automatically generate training data for a model
   • BThe ability to host a model on-premises without any cloud infrastructure
   • CThe ability to compare and experiment with multiple foundation models in a unified environment
   • DThe ability to train a custom model from scratch using your own data
   Reveal Solution Discussion Select an answer first