Prepare and test your skills

Prepare and test your skills

Free sample question with answers and detailed explanation below. The Cloud Digital Leader practice bank has an additional 3,070 practice questions.Try 10 free realistic questions: instant score, no credit card, and no account required.
A global retail company receives customer feedback in English, Spanish, and Japanese. Without training custom models or translating the feedback into a single language, the company wants to use a pre-trained Google Cloud API to automatically:
Which Google Cloud service should the company use?
Cloud Translation - Advanced API is an enterprise-grade language translation service accessed through v3 client libraries. It provides access to both the Translation LLM (TLLM) (general/translation-llm) and the Neural Machine Translation (NMT) model, along with customization, batch processing, and document formatting capabilities.
Although Cloud Translation - Advanced API supports glossaries to handle named entities and domain-specific terms consistently during translation, it does not extract those entities for analytics, evaluate customer sentiment, or classify text into topic categories. Moreover, the scenario explicitly states that the company wants to analyze the feedback without translating it into a single language.
PDF and DOCX while preserving layout, and processes large asynchronous jobs from Cloud Storage buckets.roles/cloudtranslate.user) with service accounts, supports regional endpoints, and enables billing labels.This option would be ideal if a global support team needed to translate large batches of multilingual customer cases or formatted PDF manuals into their support agents' native language while enforcing strict brand terminology using a custom glossary.
While Cloud Translation - Advanced API references "named entities" in the context of glossaries, that feature only governs how specific terms are rendered in a target language. Identifying entities and analyzing sentiment within text requires Cloud Natural Language API.
Cloud Translation - Basic API is a pre-trained translation service that converts text between more than 100 language pairs using Google's standard Neural Machine Translation (NMT) model (general/nmt). It is designed for rapid, plug-and-play integration to support multilingual communication in websites and applications.
While Cloud Translation - Basic API can detect the language of source text and convert customer feedback from Spanish or Japanese into English, it cannot understand or extract analytical insights from the text. Specifically, it does not perform entity analysis to identify products and locations, cannot conduct sentiment analysis to measure positive or negative opinions, and lacks content classification capabilities.
v2 client libraries and supports both API keys and service accounts.This service would be the right choice if the retail company needed a fast, low-configuration solution to translate casual user-generated content—such as live customer chat messages, social media posts, or product comments—from one language into another in real time.
It is a common misconception that any workload involving multiple languages requires Cloud Translation API. When the goal is natural language understanding (extracting entities, sentiment, and categories) in languages already supported by Cloud Natural Language API (such as English, Spanish, and Japanese), translation is unnecessary.
The Cloud Natural Language API is a pre-trained machine learning service that provides natural language understanding (NLU) technologies to help applications derive structured insights from unstructured text. It is designed to analyze the meaning, structure, and attitude of text across multiple supported languages without requiring custom model development.
ANALYZE_ENTITIES), the API inspects text for known entities—such as proper nouns, public figures, landmarks, and consumer goods—and returns metadata about them.ANALYZE_ENTITY_SENTIMENT) and sentiment analysis (ANALYZE_SENTIMENT), the API evaluates the prevailing emotional opinion within the text and assigns sentiment values to specific entities in English, Spanish, and Japanese.CLASSIFY_TEXT), the API analyzes text documents and assigns them to predefined content categories.language parameter is omitted.ISO-639-1 or BCP-47 identifiers when no language code is specified.ANALYZE_SYNTAX) to understand grammatical structure.ML.UNDERSTAND_TEXT function.Unlike Cloud Translation API, which converts text from a source language into a target language, Cloud Natural Language API extracts meaning, entities, sentiment, and categories directly from the text. Additionally, because it uses pre-trained models out of the box, it avoids the data preparation and training overhead required by AutoML Natural Language.
AutoML Natural Language is a Google Cloud capability that enables organizations to build and train custom machine learning models for natural language processing tasks, such as classifying content according to business-specific labels and domain-unique taxonomies.
The scenario specifies that the company wants to use a pre-trained Google Cloud API without training custom models, and only needs to categorize feedback into standard topic categories alongside entity and sentiment analysis. Implementing AutoML Natural Language requires preparing training datasets and training a custom model, which violates the requirement to avoid custom model training.
This solution would be appropriate if the retail company had a proprietary internal product catalog taxonomy and needed to train a custom machine learning model using its own labeled historical feedback data to classify text into company-specific categories.
In Google Cloud's AI strategy, organizations should first evaluate pre-trained APIs like the Cloud Natural Language API for standard use cases (such as general sentiment analysis, entity extraction, and standard content classification) and only adopt custom training solutions like AutoML Natural Language when specialized domain customization is necessary.
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