A company wants to build a chatbot that answers questions from a large corpus of legal documents. They need the answers to be grounded in the documents and include citations. They have a development team that can write code, but they want to minimize the amount of custom ML infrastructure they need to maintain. They also want to be able to customize the retrieval to prioritize certain document types. Which approach best meets these requirements?

Use Agent Platform AutoML to train a model on the legal documents

Use Model Garden to fine-tune a model on the legal documents

Use Agent Search with prebuilt RAG

Use RAG APIs to build a custom RAG pipeline with custom retrieval logic

Which developer task is most directly supported by Agent Platform?

Creating a data pipeline to ingest and clean training data

Creating a custom machine learning model from scratch using low-level TensorFlow APIs

Setting up a Kubernetes cluster to run containerized AI workloads

Building and deploying a conversational agent that can answer customer questions using company data

Which scenario is a typical use case for prebuilt RAG with Agent Search?

Building a customer support chatbot that answers questions based on the company's product documentation

Training a model to recognize objects in images

Predicting future sales based on historical data

Monitoring system logs for anomalies

A startup wants to prototype a text-generation feature. They want to try several foundation models quickly without committing to one and then integrate the chosen model into their agent. Which approach should they take?

Use Model Garden to experiment with models, then deploy via Agent Platform

Train a custom model with Agent Platform AutoML

Use RAG APIs to test model performance

Use Agent Search to generate text from indexed documents

What is the main benefit of prebuilt RAG with Agent Search?

It combines retrieval of relevant information with generation to produce grounded, context-aware responses

It eliminates the need for any data storage or indexing

It allows developers to train a custom model from scratch using their own data

It provides a fully managed service for deploying models to edge devices

Google CloudGenerative AI Leader

Domain 2Objective 4

2.4 Describe How Google Cloud Empowers Developers to Build with AI. GENERATIVE-AI-LEADER Practice Questions (Page 4)

Part of the Google Cloud’s gen AI offerings domain, which accounts for ~35% of the GENERATIVE-AI-LEADER exam.

30questions here

6free pages

7concepts

~35%of the exam

Questions 16–20

16expert · hard
A company wants to build a chatbot that answers questions from a large corpus of legal documents. They need the answers to be grounded in the documents and include citations. They have a development team that can write code, but they want to minimize the amount of custom ML infrastructure they need to maintain. They also want to be able to customize the retrieval to prioritize certain document types. Which approach best meets these requirements?
Select an answer first
17foundation · easy
Which developer task is most directly supported by Agent Platform?
Select an answer first
18foundation · easy
Which scenario is a typical use case for prebuilt RAG with Agent Search?
Select an answer first
19application · medium
A startup wants to prototype a text-generation feature. They want to try several foundation models quickly without committing to one and then integrate the chosen model into their agent. Which approach should they take?
Select an answer first
20foundation · easy
What is the main benefit of prebuilt RAG with Agent Search?
Select an answer first

Finished these 5 questions?

Review the revealed explanations, or continue through the curriculum.

Next 5 questions →Back to GENERATIVE-AI-LEADER

Free Basic Practice is a study aid with revealable answers — not a scored exam. Examers.io is independent and not affiliated with or endorsed by Google Cloud. “GENERATIVE-AI-LEADER” is a trademark of its owner, used for identification only.