Prepare and test your skills
Prepare and test your skills
Free sample question, answered and explained below. The Professional Cloud Architect practice bank has 2,335 more like it. Try 10 free: instant score, no account.
A retail enterprise is designing an automated interactive voice response (IVR) customer service system. The virtual agent must support multi-turn conversations, track session state across branching paths, extract user parameters like order numbers, and trigger backend webhooks to fulfill customer requests.
The engineering team needs to choose an architecture that minimizes custom state-machine development while delivering full conversational AI capabilities.
Which solution should the team implement?
Dialogflow CX is an enterprise-grade conversational AI platform specifically designed to orchestrate complex, multi-turn human-to-machine interactions. It provides native state-machine modeling through visual flows, pages, intent routing, condition evaluation, and integrated webhook fulfillments.

Rather than requiring engineers to manually build and maintain custom finite state machines, integrate raw speech APIs, and handle conversation context, Dialogflow CX delivers a purpose-built platform that fulfills all conversational AI requirements out of the box.
The Cloud Natural Language API provides pre-trained models for analyzing text structure, extracting document-level entities, performing sentiment analysis, classifying content into categories, and analyzing syntax (parts of speech and dependency trees).
The scenario required analyzing customer reviews, feedback emails, or support ticket transcripts for sentiment and document classification rather than running a live interactive virtual agent.
Vertex AI Search is an enterprise retrieval and discovery service that enables generative and semantic search over structured and unstructured unstructured content (such as PDFs, HTML pages, and corporate databases).
The organization needed to build an internal knowledge discovery portal or search engine allowing employees to quickly find information across internal order-processing documentation.
This approach uses the standalone Speech-to-Text API to transcribe incoming voice audio into text and the Text-to-Speech API to generate synthetic spoken replies, using Cloud Functions to execute custom business logic and Cloud Memorystore (Redis) as an external cache to store session state.
The goal was simply to perform batch audio transcription or one-off text-to-speech conversion without needing conversational natural language understanding, state tracking, or dialogue management.
Answer them yourself and get your score the moment you finish. No account needed.
Full access includes
Keep the momentum going with these hand-picked practice scenarios
Last updated