Knowledge + Practice

CCNA Gcl Google Ai Ecosystem Questions

75 of 100 questions · Page 1/2 · Gcl Google Ai Ecosystem topic · Answers revealed

Practice these questions Exam hub All questions

1

MCQeasy

A developer wants to experiment with Gemini Nano for on-device inference in a mobile app. Which Gemini API tier or environment provides access to Gemini Nano?

A.Vertex AI

B.Google AI Studio

C.Gemini API directly

D.MediaPipe and Android AICore

AnswerD

Gemini Nano is available through MediaPipe and Android AICore for on-device inference on mobile devices.

Why this answer

Gemini Nano is the smallest Gemini model designed specifically for on-device inference on mobile devices. Access to Gemini Nano is provided through MediaPipe and Android AICore, which are the frameworks that enable running the model locally on Android devices without requiring a network connection to Google's cloud servers.

Exam trap

Cisco often tests the distinction between cloud-based Gemini access (API, AI Studio, Vertex AI) and on-device inference via MediaPipe/AICore, trapping candidates who assume all Gemini models are accessed through the same cloud API.

How to eliminate wrong answers

Option A is wrong because Vertex AI is a cloud-based platform for deploying and managing machine learning models on Google Cloud, not for on-device inference on mobile apps. Option B is wrong because Google AI Studio is a web-based tool for prototyping and testing prompts with Gemini models in the cloud, not for on-device execution. Option C is wrong because the Gemini API is a cloud API that requires network connectivity to send requests to Google's servers, whereas Gemini Nano runs entirely on the device.

Practice this question →

2

Multi-Selecthard

A pharmaceutical company wants to use Google DeepMind's research breakthroughs for drug discovery. They need to identify which DeepMind technologies are directly applicable to their work. Which THREE are relevant? (Choose 3)

Select 2 answers

A.DeepMind's AI for molecular dynamics simulations

B.Gemini Ultra for general-purpose reasoning

C.Vertex AI for deploying custom models

D.AlphaFold for protein structure prediction

E.AlphaCode for generating competitive programming solutions

AnswersA, D

DeepMind has developed AI models for molecular dynamics that can accelerate drug discovery simulations.

Why this answer

AlphaFold predicts protein structures, directly applicable to drug discovery. Alphafold (not AlphaCode) is relevant. Gemini is a general-purpose model, not a specific DeepMind research tool for science.

Google Cloud offers services but not the research itself.

Practice this question →

3

MCQmedium

A media company wants to automatically generate captions for video content in multiple languages. The captions should be synced with the audio timeline. Which combination of Google Cloud services is most appropriate?

A.Cloud Video Intelligence API and Translation API

B.Document AI and Translation API

C.Speech-to-Text and Translation API

D.Text-to-Speech and Translation API

AnswerC

Speech-to-Text creates timestamped captions, Translation API translates them to other languages.

Why this answer

Option C is correct because the workflow requires first transcribing the audio track into text using Speech-to-Text API, which provides timestamps for each word or phrase to ensure synchronization with the video timeline. The resulting transcript is then passed to the Translation API to generate captions in the target languages, preserving the original timing data for alignment.

Exam trap

Cisco often tests the distinction between media analysis (Video Intelligence) and audio transcription (Speech-to-Text), leading candidates to mistakenly choose Cloud Video Intelligence API for captioning tasks that actually require speech recognition.

How to eliminate wrong answers

Option A is wrong because Cloud Video Intelligence API analyzes video content (objects, scenes, explicit content) but does not transcribe audio; it cannot generate captions from speech. Option B is wrong because Document AI is designed for extracting and processing text from documents (PDFs, invoices), not for transcribing audio or video content. Option D is wrong because Text-to-Speech API converts text into spoken audio, which is the reverse of the required workflow; it cannot transcribe existing audio into text for captioning.

Practice this question →

4

MCQhard

A financial services company needs to deploy an AI model that handles highly sensitive transaction data. They require that the model's predictions cannot be inspected by any third party, and the data must remain encrypted at all times, including during inference. Which Google Cloud feature should they use?

A.Customer-Managed Encryption Keys (CMEK)

B.Access Transparency logs

C.VPC Service Controls

D.Confidential VMs

AnswerD

Confidential VMs encrypt data in use using AMD SEV, ensuring no third party can inspect memory during inference.

Why this answer

Confidential VMs (D) are the correct choice because they provide hardware-based memory encryption using AMD Secure Encrypted Virtualization (SEV), ensuring that data remains encrypted while in use (during inference). This meets the requirement that the model's predictions cannot be inspected by any third party, including Google Cloud operators, and that data stays encrypted at all times.

Exam trap

The trap here is that candidates often confuse encryption at rest/in transit with encryption in use, and mistakenly choose CMEK or VPC Service Controls, not realizing that only Confidential VMs protect data during active computation.

How to eliminate wrong answers

Option A is wrong because Customer-Managed Encryption Keys (CMEK) protect data at rest and in transit but do not encrypt data during processing (in use), leaving it exposed in memory during inference. Option B is wrong because Access Transparency logs provide audit logs of Google Cloud administrator access but do not encrypt data or prevent third-party inspection of predictions. Option C is wrong because VPC Service Controls create a security perimeter to prevent data exfiltration but do not encrypt data in use; they control network access, not memory-level encryption.

Practice this question →

5

Multi-Selectmedium

A financial services firm must deploy a conversational AI application using Gemini on Vertex AI. They require: 1) SOC2 compliance, 2) audit logging of model inputs/outputs, and 3) data isolation within their VPC. Which THREE configurations or features should they enable?

Select 3 answers

A.Deploy the Data Loss Prevention API

B.Enable VPC Service Controls to restrict data access

C.Enable Cloud Audit Logs for Vertex AI

D.Use Google AI Studio Prototyping

E.Sign a SOC2 compliance agreement with Google Cloud

AnswersB, C, E

VPC Service Controls provide data isolation.

Why this answer

Vertex AI offers VPC Service Controls for data isolation, Cloud Audit Logs for logging, and SOC2 compliance is available. Google AI Studio does not provide these, and Data Loss Prevention API is for data masking, not a requirement.

Practice this question →

6

MCQhard

An organization needs to build an AI system that can take actions on behalf of users, such as booking appointments or sending emails. They want to use Google Cloud's capabilities for this. Which approach aligns with Google's future direction?

A.Use Gemini with function calling to execute tasks

B.Deploy a static knowledge base chatbot

C.Use AutoML Tables to predict actions

D.Fine-tune Gemini for classification

AnswerA

Function calling enables agentic behavior, allowing the model to invoke APIs to perform actions.

Why this answer

Agentic AI refers to systems that can autonomously perform tasks and take actions, which is a key area of investment.

Practice this question →

7

Multi-Selectmedium

A company is considering migrating from AWS Bedrock to Google Cloud for generative AI. They want to leverage Google's unique differentiators. Which THREE advantages does Google Cloud offer over AWS Bedrock and Azure OpenAI? (Select 3 options.)

Select 3 answers

A.Built-in model explainability for all models

B.Access to custom TPU hardware for training

C.Native multimodal support in Gemini (text, image, audio, video)

D.Integration with Google Search for grounding

E.Deep integration with Google Workspace (e.g., Gmail, Docs)

AnswersC, D, E

Gemini natively processes multiple modalities, a key differentiator.

Why this answer

Option C is correct because Gemini is natively multimodal, processing text, images, audio, and video in a single model without needing separate fine-tuned components. AWS Bedrock and Azure OpenAI require stitching together different models or services (e.g., Amazon Rekognition + Titan, or GPT-4V + Whisper) to achieve similar multimodal capabilities, which increases latency and complexity.

Exam trap

Cisco often tests the misconception that 'custom TPU hardware' is a unique differentiator for Google Cloud, but candidates forget that AWS offers its own custom chips (Trainium/Inferentia) and Azure has custom Maia chips, making this a shared capability rather than a unique advantage.

Practice this question →

8

MCQmedium

A research team is training a large multimodal model and needs to minimize training time for a fixed budget. Which Google Cloud infrastructure is specifically designed for large-scale training workloads?

A.Compute Engine with A100 GPUs

B.TPU Pods

C.Kubernetes Engine with GPU nodes

D.Cloud TPU v5e single chip

AnswerB

TPU pods are custom ASICs designed for large-scale ML training.

Why this answer

TPU pods are purpose-built for large-scale ML training, offering high-bandwidth interconnect and optimized performance for TensorFlow/JAX.

Practice this question →

9

MCQmedium

A healthcare startup needs to deploy an AI model that processes protected health information (PHI). They require a HIPAA BAA with the cloud provider. Which Google Cloud AI service can meet this compliance requirement?

A.Colab

B.Gemini API via Google AI Studio

C.Vertex AI

D.Cloud Vision API

AnswerC

Vertex AI can be covered under a HIPAA BAA.

Why this answer

Vertex AI offers a HIPAA BAA for covered entities, while other services may not.

Practice this question →

10

Multi-Selectmedium

A healthcare company needs to process medical records at scale, extracting structured information from forms and enabling natural language queries. They require HIPAA compliance and want to avoid training custom models. Which TWO services should they use? (Choose 2)

Select 2 answers

A.Natural Language AI

B.BigQuery ML

C.Document AI

D.Vertex AI Workbench

E.Vision AI

AnswersA, C

Natural Language AI can analyze unstructured medical text for entities and concepts, supporting natural language queries.

Why this answer

Natural Language AI (option A) is correct because it enables natural language queries on medical records without requiring custom model training, using pre-trained models for entity extraction, sentiment analysis, and syntax analysis. This service is HIPAA-compliant when configured appropriately, allowing healthcare companies to query structured and unstructured data at scale.

Exam trap

Cisco often tests the distinction between services that require custom model training (like BigQuery ML and Vertex AI Workbench) versus pre-built, HIPAA-compliant services (like Natural Language AI and Document AI), leading candidates to mistakenly choose options that involve training when the question explicitly prohibits it.

Practice this question →

11

MCQmedium

A company wants to build a chatbot that can answer questions about recent news events up to the current date. They want to minimize hallucinations and ensure factual accuracy. Which Google Cloud feature should they integrate?

A.Gemini API with Google Search grounding

B.Natural Language AI entity analysis

C.Document AI

D.Vertex AI Vector Search

AnswerA

Search grounding provides up-to-date information from Google Search.

Why this answer

Option A is correct because the Gemini API with Google Search grounding enables the chatbot to retrieve and cite real-time information from the web, directly addressing the need for up-to-date news and factual accuracy. Grounding connects the model's responses to verified search results, reducing hallucinations by anchoring answers in current, authoritative sources rather than relying solely on the model's training data.

Exam trap

Cisco often tests the distinction between static NLP capabilities (entity analysis, document parsing) and dynamic retrieval-augmented generation, leading candidates to choose a familiar but incorrect service like Natural Language AI or Document AI instead of recognizing that grounding requires a search-backed integration.

How to eliminate wrong answers

Option B is wrong because Natural Language AI entity analysis extracts entities (e.g., names, dates) from text but does not provide real-time grounding or fact-checking against current events; it is a static analysis tool, not a retrieval mechanism. Option C is wrong because Document AI is designed for parsing and extracting data from documents (e.g., invoices, forms), not for answering questions about live news or integrating with web search. Option D is wrong because Vertex AI Vector Search enables semantic similarity search over a fixed vector index of pre-ingested data; it cannot access or verify real-time news events unless the index is continuously updated, and it lacks the grounding mechanism to minimize hallucinations from stale or incomplete data.

Practice this question →

12

MCQhard

A machine learning engineer needs to deploy a Gemini model on an Android device for offline inference (no internet connection) to provide real-time suggestions. Which Gemini model variant is MOST appropriate?

A.Gemini Ultra

B.Gemini Pro

C.Gemini Flash

D.Gemini Nano

AnswerD

Nano is designed for on-device inference, small model size, suitable for offline usage.

Why this answer

Gemini Nano is specifically designed for on-device inference with small memory footprint. Ultra, Pro, and Flash are cloud-only models that require internet connectivity.

Practice this question →

13

Multi-Selecthard

A company is comparing Google Cloud's AI offerings with AWS Bedrock and Azure OpenAI. They need to select a platform that provides: 1) native multimodal capabilities (image, video, text in one model), 2) integration with enterprise productivity tools, and 3) custom TPU hardware for training. Which THREE differentiators are unique to Google Cloud?

Select 3 answers

A.Integration with Google Workspace (Gmail, Docs, etc.)

B.Azure OpenAI Service with GPT-4

C.Amazon Bedrock with Titan models

D.Custom TPU (Tensor Processing Unit) infrastructure for training

E.Gemini models with native multimodal understanding (text, images, video, audio)

AnswersA, D, E

Google Cloud AI integrates seamlessly with Workspace.

Why this answer

Gemini's native multimodality, Workspace integration, and TPU infrastructure are unique to Google Cloud. AWS and Azure do not offer the same combination.

Practice this question →

14

MCQeasy

A developer wants to quickly prototype a multimodal application that can process images and text using the Gemini API without incurring any cost. Which access tier should they use?

A.Colab Enterprise

B.Cloud Shell

C.Vertex AI Gemini API

D.Google AI Studio

AnswerD

Google AI Studio provides a free tier for experimentation with Gemini API.

Why this answer

Google AI Studio (option D) is the correct choice because it provides a free, browser-based environment specifically designed for prototyping with the Gemini API, including multimodal capabilities for images and text, without requiring any payment or cloud billing setup. This makes it ideal for quick experimentation and cost-free development.

Exam trap

The trap here is that candidates often confuse Google AI Studio with Vertex AI Gemini API, assuming both require billing, or they overlook that Colab Enterprise and Cloud Shell are not designed for direct, cost-free Gemini API prototyping.

How to eliminate wrong answers

Option A is wrong because Colab Enterprise is a paid Google Cloud service that requires billing activation and is intended for production-grade notebook environments, not for free prototyping. Option B is wrong because Cloud Shell provides a command-line interface and basic cloud resources but does not natively offer the Gemini API or a multimodal prototyping interface without additional setup and potential costs. Option C is wrong because the Vertex AI Gemini API is a paid enterprise service that requires a Google Cloud project with billing enabled, and while it offers the same API, it is not cost-free for prototyping.

Practice this question →

15

MCQmedium

A large enterprise is deploying a generative AI application that must meet strict latency requirements (under 500ms) and have a guaranteed uptime of 99.95% with full audit logging. Which Google Cloud environment should they use for the Gemini API?

A.Google AI Studio (free tier)

B.Cloud Functions with a direct API call

C.Gemini API through a public endpoint without any project

D.Vertex AI

AnswerD

Vertex AI provides VPC controls, SLA guarantees, audit logging, and compliance certifications — suitable for enterprise production deployments.

Why this answer

Vertex AI is the correct choice because it provides enterprise-grade features required for this deployment: a guaranteed 99.95% uptime SLA, full audit logging via Cloud Audit Logs, and the ability to meet sub-500ms latency through regional endpoints and optimized infrastructure. Google AI Studio and public endpoints lack SLAs and audit logging, while Cloud Functions adds unnecessary compute overhead and does not inherently provide the required uptime guarantee.

Exam trap

The trap here is that candidates may assume any API call can meet enterprise requirements, but Cisco tests the understanding that only Vertex AI provides the SLA, audit logging, and latency optimization needed for production generative AI workloads.

How to eliminate wrong answers

Option A is wrong because Google AI Studio (free tier) has no SLA, no audit logging, and is intended for prototyping only, not production workloads with strict latency and uptime requirements. Option B is wrong because Cloud Functions with a direct API call introduces additional cold-start latency and does not provide a guaranteed 99.95% uptime SLA or built-in audit logging for the Gemini API itself. Option C is wrong because a public endpoint without a project lacks any SLA, audit logging, and rate limiting, making it unsuitable for enterprise production use.

Practice this question →

16

MCQmedium

A company is evaluating Google Cloud vs AWS Bedrock for building a multimodal application that needs to understand images, video, and text in a single model. Which unique Google Cloud capability supports this requirement?

A.Gemini API's native multimodal support across text, images, video, and audio

B.Use of Hugging Face models on Vertex AI

C.Integration with Amazon Bedrock's Titan model

D.BigQuery ML's image analysis functions

AnswerA

Gemini is designed for multimodal inputs in a single model.

Why this answer

Gemini is a natively multimodal model capable of processing text, images, audio, video, and code. AWS Bedrock's models are primarily text or image-only, lacking native video understanding.

Practice this question →

17

Multi-Selecteasy

A startup is building a multimodal application that needs to process both images and text. They want to prototype quickly for free before moving to production with enterprise controls. Which TWO services should they use? (Choose 2)

Select 2 answers

A.Google AI Studio for prototyping

B.Vertex AI for prototyping

C.Google Colab for production

D.Vision AI and Natural Language AI separately

E.Vertex AI for production deployment

AnswersA, E

AI Studio provides free access to Gemini models for rapid prototyping of multimodal applications.

Why this answer

Google AI Studio is correct because it provides a free, browser-based environment for prototyping multimodal applications that process both images and text, using models like Gemini. It allows rapid experimentation without cost or infrastructure setup, making it ideal for quick prototyping before moving to production.

Exam trap

Cisco often tests the distinction between prototyping and production services, trapping candidates who confuse Vertex AI (production) with AI Studio (prototyping) or who think Colab is suitable for production deployment.

Practice this question →

18

MCQmedium

A company using Vertex AI for model training wants to reduce costs by automatically stopping training when the model's performance stops improving. Which Vertex AI feature should they configure?

A.Early stopping (via Vertex AI Training)

B.AutoML

C.Hyperparameter tuning

D.Vertex AI Workbench

AnswerA

Early stops training when a metric stops improving, saving time and cost.

Why this answer

Early stopping in Vertex AI Training stops training runs when performance plateaus. Hyperparameter tuning optimizes parameters, not cost. Workbench is for notebooks.

AutoML handles this automatically but is for specific model types.

Practice this question →

19

MCQeasy

A developer wants to use a pre-trained model to identify objects in images. Which Google Cloud AI API should they use?

A.Speech-to-Text

B.Translation API

C.Natural Language AI

D.Vision AI

AnswerD

Vision AI offers pre-trained models for object detection in images.

Why this answer

Vision AI provides pre-trained models for object detection, image classification, etc. Natural Language AI is for text, Speech-to-Text for audio, and Translation for text translation.

Practice this question →

20

MCQhard

A financial services firm needs to use Gemini for analyzing customer transaction data. They require that all data remain within their VPC and that model inference logs be auditable. Which access tier should they choose?

A.Colab Enterprise

B.Gemini API without Vertex AI

C.Vertex AI

D.Google AI Studio

AnswerC

Vertex AI offers VPC Service Controls, data isolation, and audit logging for enterprise compliance.

Why this answer

Vertex AI provides enterprise controls like VPC-SC, data isolation, and audit logging, while Google AI Studio is a prototyping environment without these guarantees.

Practice this question →

21

Multi-Selectmedium

A data science team is choosing between Gemini Pro and Gemini Flash for a real-time content moderation API. They need low cost and low latency, but can tolerate slightly lower accuracy. Which TWO Gemini variants should they consider?

Select 2 answers

A.Gemini Nano

B.Gemini Flash

C.Gemini 1.5 Pro

D.Gemini Ultra

E.Gemini Pro

AnswersB, E

Gemini Flash is designed for low latency and low cost, fitting the primary requirements.

Why this answer

Gemini Flash is designed for low latency and cost efficiency, making it ideal for real-time applications like content moderation where speed and budget are prioritized over peak accuracy. Gemini Pro offers a balanced trade-off with higher accuracy than Flash but still lower latency and cost than larger models, fitting scenarios where slight accuracy loss is acceptable but better performance than Flash is desired.

Exam trap

Cisco often tests the distinction between model variants (Nano, Flash, Pro, Ultra) and their specific trade-offs, trapping candidates who confuse 'lowest cost' with 'best for real-time' or who overlook that Gemini Pro is a valid option for slightly lower accuracy tolerance, not just Flash.

Practice this question →

22

Multi-Selecteasy

A company wants to build a multimodal application that can analyze images and generate captions. They are considering using Google Cloud AI. Which TWO services can be directly used for this purpose?

Select 2 answers

A.Vision AI

B.Text-to-Speech AI

C.Natural Language AI

D.Translation AI

E.Gemini API

AnswersA, E

Vision AI can analyze images to extract features and labels.

Why this answer

Vision AI (option A) is correct because it provides pre-trained models for image analysis, including object detection and image labeling, which are essential for generating captions. The Gemini API (option E) is correct because it is a multimodal model that can directly process images and text, enabling it to analyze images and produce descriptive captions without needing separate services.

Exam trap

The trap here is that candidates may mistakenly think Natural Language AI (option C) can generate captions because it handles text, but it cannot process images, while the Gemini API's multimodal capability is often overlooked in favor of more familiar single-purpose services.

Practice this question →

23

MCQmedium

A company wants to extract structured data from scanned invoices. Which Google Cloud AI service is purpose-built for this use case?

A.Vision AI

B.Translation API

C.Natural Language AI

D.Document AI

AnswerD

Document AI specializes in processing scanned documents and extracting structured data.

Why this answer

Document AI is designed for document understanding, including invoice parsing.

Practice this question →

24

MCQeasy

A data science team wants to collaborate on a machine learning project using a shared notebook environment that supports Python and R, with free access to GPUs for prototyping. Which Google tool is most suitable?

A.Kaggle

B.Cloud Shell

C.Vertex AI Workbench

D.Colab

AnswerD

Colab offers free notebooks with GPU support, ideal for collaborative prototyping.

Why this answer

Colab (D) is the most suitable Google tool because it provides a shared notebook environment that supports both Python and R via the `%load_ext rpy2.ipython` magic, offers free access to GPUs (e.g., NVIDIA T4 or V100) for prototyping, and enables real-time collaboration through Google Drive integration. This directly matches the team's requirements for a collaborative, multi-language notebook with free GPU resources.

Exam trap

Cisco often tests the distinction between free-tier prototyping tools (Colab) and enterprise-managed services (Vertex AI Workbench), where candidates mistakenly choose Vertex AI Workbench because it supports both Python and R, overlooking the explicit requirement for 'free access to GPUs' which only Colab provides in this context.

How to eliminate wrong answers

Option A is wrong because Kaggle is a data science competition platform that, while offering free GPU access and notebooks, does not natively support R in its notebook environment (Kaggle Notebooks are Python-only) and is not a Google-owned tool for general collaboration—it is owned by Google but primarily designed for competitions, not team-based project prototyping. Option B is wrong because Cloud Shell is a browser-based terminal with a temporary environment (5 GB persistent disk, 2 vCPUs) that lacks a notebook interface, does not support R, and provides no GPU access—it is meant for command-line operations and small-scale testing, not ML prototyping. Option C is wrong because Vertex AI Workbench is a fully managed Jupyter notebook service that supports Python and R, but it requires a Google Cloud project with billing enabled, does not offer free GPU access (GPUs incur costs), and is designed for production workloads, not free prototyping.

Practice this question →

25

MCQeasy

A data analyst wants to train and deploy a custom image classification model with minimal ML engineering overhead. Which Google Cloud service should they use?

A.Google Colab

B.Vertex AI

C.AI APIs (Vision AI)

D.BigQuery ML

AnswerB

Vertex AI offers end-to-end ML workflows including AutoML for image classification, custom training, and deployment with managed infrastructure.

Why this answer

Vertex AI is the correct choice because it provides a fully managed, end-to-end platform for training and deploying custom machine learning models, including image classification, with minimal ML engineering overhead. It offers AutoML capabilities for automated model training, managed pipelines, and one-click deployment to a scalable endpoint, eliminating the need for infrastructure management.

Exam trap

Cisco often tests the distinction between using pre-built AI APIs (like Vision AI) for standard tasks versus Vertex AI for custom model training, leading candidates to mistakenly choose Vision AI when the question explicitly requires a custom model.

How to eliminate wrong answers

Option A is wrong because Google Colab is a Jupyter notebook environment for prototyping and experimentation, not a production-grade service for training and deploying custom models with minimal overhead; it lacks managed deployment and scaling capabilities. Option C is wrong because AI APIs (Vision AI) are pre-trained, general-purpose models for common tasks like object detection or OCR, not for training custom image classification models from scratch. Option D is wrong because BigQuery ML is designed for creating and executing machine learning models directly on structured data in BigQuery using SQL, not for custom image classification tasks that require image data and deep learning frameworks.

Practice this question →

26

Multi-Selecthard

A financial institution needs to deploy a document processing pipeline to extract data from invoices. They must ensure data isolation, audit logging, and compliance with SOC 2. Which THREE Google Cloud services should they combine?

Select 3 answers

A.Document AI

B.Cloud Vision API

C.Cloud Audit Logs

D.Vertex AI

E.BigQuery

AnswersA, C, D

Document AI extracts data from invoices.

Why this answer

Document AI (Option A) is correct because it is purpose-built for extracting structured data from documents like invoices using pre-trained models (e.g., the Invoice Parser), enabling accurate key-value pair extraction. Cloud Audit Logs (Option C) is correct because it provides immutable, tamper-evident logs of all API calls and administrative actions, which is essential for SOC 2 compliance and audit trails. Vertex AI (Option D) is correct because it can host custom machine learning models for post-processing or validation of extracted data, and its integration with Cloud Audit Logs ensures data isolation and governance within a single tenant project.

Exam trap

The trap here is that candidates often confuse Cloud Vision API with Document AI, assuming OCR alone is sufficient for structured data extraction, but Document AI's specialized parsers and audit logging integration are required for compliance and accuracy in financial document processing.

Practice this question →

27

MCQeasy

Which Google Cloud AI API is used to convert spoken language into text?

A.Speech-to-Text API

B.Text-to-Speech API

C.Natural Language API

D.Translation API

AnswerA

Speech-to-Text transcribes audio to text.

Why this answer

Speech-to-Text API transcribes audio into text.

Practice this question →

28

Multi-Selecthard

A healthcare company wants to use Google Cloud AI services for analyzing medical images and patient records. They need to ensure data is not used to improve Google's models and that all model predictions are logged for audit. Which TWO configurations must be implemented? (Choose TWO)

Select 2 answers

A.Enable Access Transparency logs

B.Use a shared VPC network

C.Use VPC Service Controls to restrict data access

D.Enable Customer-Managed Encryption Keys (CMEK) for the AI services

E.Disable data logging in the AI services' project-level settings

AnswersD, E

CMEK gives the customer control over encryption keys, an additional security measure, and is often required for compliance.

Why this answer

Option D is correct because Customer-Managed Encryption Keys (CMEK) allow the healthcare company to control the encryption keys used to protect their data at rest in Google Cloud AI services. This ensures that Google cannot access the data for model improvement, as the data remains encrypted under keys managed by the customer. Option E is correct because disabling data logging in the AI services' project-level settings prevents the service from collecting and storing input data or predictions, which is a direct way to ensure data is not used to improve Google's models and that audit requirements are met by controlling what is logged.

Exam trap

The trap here is that candidates often confuse VPC Service Controls (which prevent data exfiltration) with data usage controls for model training, or they think Access Transparency logs are sufficient for audit logging of predictions, when in fact they only log administrative actions, not model outputs.

Practice this question →

29

MCQhard

A company is comparing Google Cloud Vertex AI with AWS Bedrock and Azure OpenAI. They need a model that can natively process text, images, audio, and video. Which differentiator does Google Cloud offer?

A.Gemini's native multimodal capabilities

B.Access to GPT-4o

C.Support for Anthropic Claude

D.Integration with Microsoft Copilot

AnswerA

Gemini is designed from the ground up as multimodal.

Why this answer

Gemini is Google's multimodal model natively trained on text, images, audio, and video from the ground up, enabling it to process and reason across these modalities without separate components. This native capability is a key differentiator for Google Cloud Vertex AI, as competing platforms like AWS Bedrock and Azure OpenAI primarily offer models that are text-centric or require separate models for different modalities.

Exam trap

Cisco often tests the distinction between 'native multimodal' models (trained on all modalities simultaneously) versus 'composite multimodal' systems that combine separate models for each modality, leading candidates to overestimate the capabilities of GPT-4o or Claude.

How to eliminate wrong answers

Option B is wrong because GPT-4o, while capable of processing text, images, and audio, is not natively trained on video; it processes video as a sequence of frames, not as a unified video modality. Option C is wrong because Anthropic Claude is primarily a text-based model with limited image understanding, lacking native support for audio and video processing. Option D is wrong because Microsoft Copilot is an integration layer that leverages underlying models (like GPT-4) and does not itself provide native multimodal capabilities across text, images, audio, and video.

Practice this question →

30

MCQmedium

A research team wants to run a large-scale training job for a custom transformer model. They need access to Google's custom AI accelerators with high-speed interconnects for distributed training. Which infrastructure should they use?

A.Vertex AI Training with default GPU

B.Cloud TPU v3-32 pod

C.Compute Engine with NVIDIA H100 GPUs

D.Google Kubernetes Engine with GPU nodes

AnswerB

TPU pods provide custom accelerators with high-speed interconnects.

Why this answer

The correct answer is Cloud TPU v3-32 pod because it provides Google's custom AI accelerators (TPUs) with high-speed interconnects (e.g., a 2D toroidal mesh network) specifically designed for large-scale distributed training of transformer models. This pod offers 32 TPU v3 chips interconnected at 100 Gbps per chip, enabling efficient model parallelism and data parallelism for custom transformer architectures, which is not achievable with standard GPUs or Kubernetes setups.

Exam trap

The trap here is that candidates may confuse 'custom AI accelerators' with any high-end GPU (like H100) or assume Kubernetes provides equivalent distributed training performance, but the question specifically requires Google's custom TPUs with high-speed interconnects, which only the TPU pod offers.

How to eliminate wrong answers

Option A is wrong because Vertex AI Training with default GPU uses standard NVIDIA GPUs (e.g., T4 or V100) without Google's custom TPU accelerators or the high-speed interconnects required for large-scale distributed training; it is designed for smaller-scale jobs. Option C is wrong because Compute Engine with NVIDIA H100 GPUs, while powerful, lacks the custom TPU architecture and the dedicated high-speed interconnects (e.g., TPU pod's 2D torus) that are optimized for synchronous distributed training of large transformer models; H100s use NVLink/NVSwitch which are less efficient for the specific scale and topology needed. Option D is wrong because Google Kubernetes Engine with GPU nodes relies on standard GPU hardware and Kubernetes networking (e.g., VPC, Calico) which introduces latency and bandwidth bottlenecks compared to the TPU pod's dedicated interconnects, and it does not provide Google's custom TPU accelerators.

Practice this question →

31

Multi-Selecteasy

A data scientist wants to run ML models directly on their BigQuery data without moving data out. Which THREE statements about BigQuery ML are correct? (Choose 3)

Select 3 answers

A.BigQuery ML supports binary logistic regression models

B.BigQuery ML supports deep neural network models like CNNs and RNNs

C.BigQuery ML requires data to be exported to Cloud Storage before training

D.BigQuery ML supports creating and evaluating linear regression models using SQL

E.BigQuery ML can be used for recommendation systems using matrix factorization

AnswersA, D, E

Binary logistic regression is supported for classification tasks using SQL.

Why this answer

BigQuery ML supports binary logistic regression models because it provides the `CREATE MODEL` statement with `OPTIONS(model_type='LOGISTIC_REG')` for classification tasks. This allows data scientists to train and evaluate models directly on data stored in BigQuery using standard SQL syntax, without needing to export or move the data.

Exam trap

Cisco often tests the misconception that BigQuery ML supports all types of ML models, including deep neural networks, when in fact it only supports a curated set of simpler, SQL-friendly model types like linear and logistic regression, matrix factorization, and boosted trees.

Practice this question →

32

MCQhard

A company is deploying a generative AI application that must meet HIPAA compliance. They need audit logging, data isolation, and a Business Associate Agreement (BAA). Which Gemini API access tier should they use?

A.Vertex AI

B.Google AI Studio

C.Gemini Nano on-device

D.Cloud Vision API

AnswerA

Vertex AI provides VPC Service Controls, CMEK, audit logging, and supports HIPAA BAA, making it suitable for regulated healthcare workloads.

Why this answer

Vertex AI is the correct choice because it is the only Google Cloud platform that offers HIPAA-compliant services with support for audit logging, data isolation, and the ability to sign a Business Associate Agreement (BAA). Vertex AI provides enterprise-grade security controls, including VPC Service Controls, CMEK, and access transparency, which are required for protected health information (PHI) under HIPAA. The Gemini API accessed through Vertex AI inherits these compliance capabilities, making it suitable for regulated healthcare workloads.

Exam trap

Cisco often tests the misconception that Google AI Studio is a production-ready tier for compliance workloads, but it lacks BAA support and enterprise controls, making Vertex AI the only valid option for HIPAA.

How to eliminate wrong answers

Option B is wrong because Google AI Studio is a free, web-based prototyping tool that does not support HIPAA compliance, BAA signing, or enterprise audit logging; it is designed for experimentation, not production use with PHI. Option C is wrong because Gemini Nano on-device runs locally on a device and lacks centralized audit logging, data isolation controls, and the infrastructure to support a BAA, making it unsuitable for HIPAA compliance. Option D is wrong because Cloud Vision API is a separate service for image analysis, not a Gemini API access tier, and while it can be HIPAA-compliant when used with Vertex AI, it does not provide the generative AI capabilities of Gemini.

Practice this question →

33

MCQhard

A developer is building a mobile app that needs to run an AI model on-device for low-latency inference even without internet. Which Gemini model variant is designed for on-device deployment?

A.Gemini Pro

B.Gemini Ultra

C.Gemini Flash

D.Gemini Nano

AnswerD

Nano is designed for on-device inference.

Why this answer

Gemini Nano is optimized for on-device execution on mobile devices.

Practice this question →

34

MCQeasy

A researcher wants to quickly prototype an AI model using a free, hosted Jupyter notebook environment with GPU support. Which Google product is BEST suited for this?

A.AI Platform Notebooks

B.Kaggle Notebooks

C.Google Colab

D.Vertex AI Workbench

AnswerC

Colab offers free, hosted Jupyter notebooks with GPU, ideal for quick prototyping.

Why this answer

Google Colab provides free hosted Jupyter notebooks with GPU support. Kaggle is for datasets and competitions but not primarily for notebooks. Vertex AI Workbench and AI Platform are enterprise tools with costs.

Practice this question →

35

MCQhard

A data science team wants to run a large-scale transformer training job with custom model architectures. They need the highest compute density for a multi-node job and want to minimize inter-node communication latency. Which Google Cloud infrastructure is BEST suited for this workload?

A.A single TPU v5e VM with multiple accelerators

B.A cluster of A100 GPU VMs connected via standard networking

C.TPU v4 Pod slices with high-speed inter-chip interconnect

D.Cloud Run jobs with GPU acceleration

AnswerC

TPU v4 Pods provide massive compute density and fast inter-chip communication, optimal for large transformer training.

Why this answer

TPU v4 Pod slices provide the highest compute density for multi-node transformer training by using a custom 3D torus interconnect with 800 Gbps per chip bandwidth, which minimizes inter-node communication latency far below what standard networking can achieve. This architecture is specifically designed for large-scale model parallelism, making it ideal for custom transformer architectures that require frequent all-reduce and collective communication operations.

Exam trap

The trap here is that candidates often assume GPU clusters with standard networking (Option B) are sufficient for multi-node training, underestimating how drastically inter-node latency impacts scaling efficiency for transformer models with large parameter counts.

How to eliminate wrong answers

Option A is wrong because a single TPU v5e VM, while powerful, is limited to a single host and cannot scale to multi-node training without incurring network latency between separate TPU hosts. Option B is wrong because A100 GPU VMs connected via standard networking (e.g., 100 Gbps Ethernet) introduce significant inter-node latency compared to TPU Pods' dedicated high-speed interconnects, which are optimized for synchronous training. Option D is wrong because Cloud Run jobs are serverless and designed for stateless, short-lived tasks, not for large-scale, multi-node distributed training with custom model architectures that require persistent, high-bandwidth inter-node communication.

Practice this question →

36

MCQmedium

A company wants to ground Gemini responses with real-time Google Search results to improve accuracy of current events. Which feature enables this?

A.BigQuery ML

B.Vertex AI Vector Search

C.Gemini API function calling

D.Google Search grounding

AnswerD

This feature links Gemini to Google Search.

Why this answer

Google Search grounding connects Gemini to live search results to reduce hallucinations and provide up-to-date information.

Practice this question →

37

MCQmedium

A data analyst wants to experiment with ML models using a free, cloud-based Jupyter notebook environment with GPU support. Which Google tool should they use?

A.Cloud Shell

B.Google Colab

C.BigQuery Studio

D.Vertex AI Workbench

AnswerB

Google Colab offers free notebooks with GPU support for ML experimentation.

Why this answer

Google Colab is a free, cloud-based Jupyter notebook environment that provides access to GPUs (e.g., NVIDIA T4 or V100) without requiring any setup or billing. It is specifically designed for experimentation and learning with ML models, making it the correct choice for a data analyst seeking a free, GPU-enabled notebook environment.

Exam trap

The trap here is that candidates may confuse Vertex AI Workbench (a paid, enterprise tool) with a free offering because both provide Jupyter notebooks with GPU support, but the question's requirement for a 'free' tool eliminates it.

How to eliminate wrong answers

Option A is wrong because Cloud Shell is a browser-based terminal with a small, ephemeral VM (typically 5 GB of persistent disk) that does not include GPU support or a Jupyter notebook interface; it is meant for command-line operations and not for running ML experiments. Option C is wrong because BigQuery Studio is a data analysis and SQL workspace within BigQuery that focuses on querying large datasets, not on running ML model training with GPU acceleration in a notebook environment. Option D is wrong because Vertex AI Workbench is a fully managed, enterprise-grade Jupyter notebook environment that does support GPUs, but it requires a paid Google Cloud project and is not free; the question explicitly asks for a free tool.

Practice this question →

38

MCQmedium

A data scientist wants to share a Jupyter notebook with interactive visualizations for a Kaggle competition. Which tool should they use to easily publish and collaborate?

A.Google Colab

B.Kaggle Notebooks

C.Vertex AI Workbench

D.Cloud Shell Editor

AnswerB

Kaggle Notebooks are integrated with Kaggle datasets and competitions.

Why this answer

B is correct because Kaggle Notebooks are natively integrated with the Kaggle competition platform, allowing data scientists to publish interactive visualizations directly within the competition environment and collaborate with other participants. Unlike generic tools, Kaggle Notebooks provide seamless access to competition datasets, GPU/TPU resources, and version control tailored for Kaggle workflows.

Exam trap

Cisco often tests the distinction between general-purpose tools (like Colab) and platform-specific tools (like Kaggle Notebooks), trapping candidates who overlook the tight integration with competition datasets and submission APIs.

How to eliminate wrong answers

Option A is wrong because Google Colab is a general-purpose cloud-based Jupyter notebook environment that lacks native integration with Kaggle competition datasets and submission systems, requiring manual uploads and API workarounds. Option C is wrong because Vertex AI Workbench is an enterprise MLOps tool designed for production model development and deployment, not for lightweight sharing and collaboration on Kaggle competitions. Option D is wrong because Cloud Shell Editor is a lightweight code editor in the Google Cloud console intended for quick edits and terminal access, not for publishing interactive Jupyter notebooks with visualizations.

Practice this question →

39

Multi-Selectmedium

A company is choosing between Gemini Pro and Gemini Ultra for a document summarization task. Which THREE factors should they consider when deciding between the two model variants? (Choose THREE)

Select 3 answers

A.Model capability and accuracy

B.Availability in Google AI Studio

C.Latency requirements

D.Cost per token

E.Multimodal support

AnswersA, C, D

Ultra is more capable and accurate for complex tasks, but Pro may suffice for simpler summarization.

Why this answer

Model capability and accuracy (A) is correct because Gemini Pro is optimized for high-throughput, cost-efficient tasks with lower accuracy demands, while Gemini Ultra is designed for complex, high-accuracy reasoning. The choice directly impacts the quality of the summarization output, as Ultra uses a larger parameter count and more advanced attention mechanisms to handle nuanced context, whereas Pro may struggle with ambiguous or lengthy documents.

Exam trap

Cisco often tests the misconception that model availability or multimodal support are primary selection criteria, when in fact the core trade-offs are capability, latency, and cost per token for the specific task.

Practice this question →

40

MCQmedium

A financial analyst wants to run a regression model inside BigQuery using SQL, without moving data to a separate ML environment. Which Google Cloud service allows this directly?

A.BigQuery ML

B.AI Platform

C.AutoML Tables

D.Vertex AI

AnswerA

BigQuery ML allows creating and running ML models using SQL directly in BigQuery.

Why this answer

BigQuery ML enables creating and running ML models using standard SQL queries directly in BigQuery. Vertex AI and AI Platform require data to be moved, and AutoML is a higher-level service but still not SQL-based.

Practice this question →

41

MCQeasy

A startup wants to quickly prototype a multimodal AI application that can process images and text using Gemini. They have minimal budget and need a free tier for initial development. Which access tier should they use?

A.Vertex AI

B.Google AI Studio

C.Cloud TPU

D.TensorFlow Hub

AnswerB

Google AI Studio offers a free tier with rate limits, perfect for experimentation and prototyping before moving to production.

Why this answer

Google AI Studio is the correct choice because it offers a free tier specifically designed for rapid prototyping with Gemini models, including multimodal capabilities for processing images and text. It provides a web-based interface and API access without requiring a billing account, making it ideal for startups with minimal budget. Vertex AI, while powerful, requires a paid Google Cloud project and is intended for production deployment, not free-tier prototyping.

Exam trap

The trap here is that candidates often confuse Vertex AI's free trial credits (which still require a billing account) with a true free tier, or they assume TensorFlow Hub provides API access to Gemini, when in fact it only hosts static model artifacts for download.

How to eliminate wrong answers

Option A is wrong because Vertex AI is a managed ML platform that requires a billing-enabled Google Cloud project and charges for usage, making it unsuitable for a minimal-budget free-tier prototype. Option C is wrong because Cloud TPU is a hardware accelerator for training large models, not an access tier for using pre-built multimodal APIs like Gemini, and it incurs significant costs. Option D is wrong because TensorFlow Hub is a repository for reusable model components, not a service for accessing Gemini's multimodal capabilities, and it does not provide a free tier for API-based prototyping.

Practice this question →

42

MCQhard

A data scientist wants to train a large transformer model from scratch on custom data. They anticipate the training will require thousands of TPU-v5e chips for several weeks. Which Google Cloud infrastructure component is designed for this scale?

A.TPU v5e Pod slice (e.g., 256-chip pod)

B.BigQuery ML

C.Single TPU v5e device

D.Compute Engine with A100 GPU clusters

AnswerA

TPU v5e Pods are designed for large-scale distributed training with fast inter-chip interconnects.

Why this answer

A is correct because a TPU v5e Pod slice (e.g., a 256-chip pod) is specifically designed for large-scale distributed training, providing high-bandwidth inter-chip interconnect (ICI) and collective communication optimizations that enable efficient scaling across thousands of chips for weeks-long training runs. This infrastructure component is purpose-built for the massive parallelism and fault tolerance required when training large transformer models from scratch on custom data.

Exam trap

The trap here is that candidates may confuse a single TPU device (Option C) with a Pod slice, not realizing that 'thousands of chips for several weeks' explicitly requires the multi-chip pod architecture with high-speed interconnects, not just a single accelerator.

How to eliminate wrong answers

Option B is wrong because BigQuery ML is a serverless machine learning service for running SQL-based ML models on structured data in BigQuery, not a hardware infrastructure component designed for large-scale distributed training of transformer models. Option C is wrong because a single TPU v5e device lacks the memory capacity, compute power, and inter-chip connectivity needed to train a large transformer model requiring thousands of chips over several weeks. Option D is wrong because while Compute Engine with A100 GPU clusters can be used for training, the question specifically asks for the Google Cloud infrastructure component designed for this scale, and TPU v5e Pod slices are Google's optimized solution for large-scale transformer training, offering superior inter-chip bandwidth and scaling efficiency compared to GPU clusters.

Practice this question →

43

MCQmedium

A company wants to use Google DeepMind's advances in protein structure prediction to accelerate drug discovery. Which DeepMind achievement is most relevant to this goal?

A.AlphaFold

B.WaveNet

C.AlphaCode

D.AlphaGo

AnswerA

AlphaFold predicts protein structures, directly applicable to drug discovery.

Why this answer

AlphaFold solves protein structure prediction, a key enabler for drug discovery. AlphaGo is for board games, AlphaCode for programming, and WaveNet for audio.

Practice this question →

44

MCQhard

A healthcare organization needs to process sensitive patient data (PHI) using Google Cloud's Natural Language API for entity extraction. They require HIPAA compliance and data isolation within Google's infrastructure. What is the minimum configuration they must enable?

A.Sign a HIPAA Business Associate Agreement (BAA) with Google Cloud and enable VPC Service Controls

B.Use the Healthcare API instead of Natural Language API

C.Enable CMEK for the Natural Language API and use a private IP address

D.Deploy the Natural Language API in a Google-managed project and use a shared VPC

AnswerA

A signed BAA is mandatory for HIPAA compliance, and VPC Service Controls provide data isolation and prevent data exfiltration.

Why this answer

To use Cloud AI APIs with HIPAA compliance and data isolation, the customer must sign a BAA and use VPC Service Controls to prevent data exfiltration.

Practice this question →

45

Multi-Selecthard

A research lab is planning to train a massive protein folding model similar to AlphaFold. They want to use Google Cloud infrastructure and tools. Which THREE components are most relevant?

Select 3 answers

A.Cloud TPU pods

B.Cloud Vision API

C.Vertex AI Pipeline

D.Google AI Studio

E.Google DeepMind collaboration

AnswersA, C, E

TPU pods are ideal for large-scale training of scientific models.

Why this answer

Cloud TPU pods are specifically designed for large-scale machine learning workloads like protein folding, offering high-throughput matrix operations essential for training models similar to AlphaFold. They provide the massive parallel compute power needed for training deep neural networks on protein structure prediction tasks, which require processing large datasets and complex 3D spatial relationships.

Exam trap

The trap here is that candidates may confuse Google's pre-built AI services (like Vision API or AI Studio) with the specialized infrastructure needed for training custom large-scale models, overlooking that TPU pods are the core compute resource for such workloads.

Practice this question →

46

Multi-Selecthard

A company is evaluating Google Cloud's AI portfolio versus competitors. They want to leverage Gemini's unique capabilities. Which THREE differentiators should they highlight when comparing to AWS Bedrock and Azure OpenAI? (Choose 3)

Select 3 answers

A.Vertex AI as a unified ML platform

B.Custom TPU hardware for training large models

C.Grounding with Google Search for real-time, verifiable responses

D.Integration with Google Workspace (e.g., Gmail, Docs)

E.Native multimodal understanding across text, images, video, and audio

AnswersC, D, E

Google Cloud offers native grounding with Google Search, reducing hallucinations and providing citations; AWS/Azure have similar features but not with Google Search.

Why this answer

Option C is correct because Gemini's grounding with Google Search allows it to access and cite real-time, verifiable information from the web, reducing hallucinations and improving factual accuracy. This is a unique differentiator as AWS Bedrock and Azure OpenAI do not natively integrate a live search engine for grounding responses, requiring custom RAG implementations instead.

Exam trap

Cisco often tests the distinction between platform-level features (like Vertex AI or TPUs) and model-level differentiators (like grounding or multimodal understanding), causing candidates to select options that are true for Google Cloud but not unique to Gemini.

Practice this question →

47

Multi-Selecteasy

Which TWO of the following are valid Google Cloud AI APIs for natural language processing?

Select 2 answers

A.Speech-to-Text API

B.Translation API

C.Vision AI

D.Natural Language API

E.Document AI

AnswersB, D

Translation API translates text between languages, a core NLP task.

Why this answer

The Translation API and Natural Language API are both valid Google Cloud AI APIs for natural language processing (NLP). The Translation API converts text between languages, which is a core NLP task, while the Natural Language API provides entity recognition, sentiment analysis, and syntax analysis, directly processing human language.

Exam trap

Cisco often tests the distinction between APIs that process language (NLP) versus those that process speech or images, leading candidates to mistakenly include Speech-to-Text or Vision AI as NLP APIs when they are actually in different AI domains.

Practice this question →

48

MCQmedium

A company has a large dataset of customer support tickets stored in BigQuery. They want to predict ticket severity (high, medium, low) using SQL queries without moving data out of BigQuery. Which service should they use?

A.Vertex AI AutoML Natural Language

B.Cloud Natural Language API

C.Gemini API

D.BigQuery ML

AnswerD

BigQuery ML enables training and prediction using SQL on data already in BigQuery.

Why this answer

BigQuery ML (D) is correct because it allows users to create and execute machine learning models directly within BigQuery using SQL, without moving data out of the warehouse. For a classification task like predicting ticket severity, BigQuery ML supports models such as logistic regression, boosted trees, and deep neural networks, all trained and deployed using standard SQL queries on data already in BigQuery.

Exam trap

Cisco often tests the distinction between services that require data movement (like Vertex AI AutoML) versus services that operate directly on the data warehouse (like BigQuery ML), and the trap here is assuming that any ML service in Google Cloud must involve exporting data to a separate AI platform.

How to eliminate wrong answers

Option A is wrong because Vertex AI AutoML Natural Language requires exporting data from BigQuery to a Cloud Storage bucket and then importing it into Vertex AI, which violates the requirement of not moving data out of BigQuery. Option B is wrong because Cloud Natural Language API is a pre-trained API for sentiment analysis, entity extraction, and syntax analysis, not a custom classification model that can be trained on the company's specific ticket severity labels. Option C is wrong because Gemini API is a generative AI API for tasks like text generation and summarization, not designed for custom classification model training or SQL-based ML workflows within BigQuery.

Practice this question →

49

Multi-Selectmedium

A financial services company needs to deploy an AI model for fraud detection. They require that the model's predictions be explainable, the data never leaves their VPC, and they need a guarantee of 99.95% uptime. Which TWO Google Cloud offerings should they consider? (Select 2 options.)

Select 2 answers

A.Vertex AI (with VPC Service Controls and SLA)

B.BigQuery ML

C.Vertex AI Explainable AI

D.Cloud Functions

E.AI Platform Prediction

AnswersA, C

Vertex AI supports VPC Service Controls for data isolation and offers 99.95% uptime SLA.

Why this answer

Vertex AI (with VPC Service Controls and SLA) is correct because it allows the company to deploy the fraud detection model within their VPC using VPC Service Controls, ensuring data never leaves their network, and the SLA provides a 99.95% uptime guarantee. Vertex AI Explainable AI is correct because it provides built-in feature attributions and explanations for model predictions, meeting the explainability requirement for fraud detection models.

Exam trap

Cisco often tests the distinction between legacy services (AI Platform Prediction) and their modern replacements (Vertex AI), and candidates may confuse BigQuery ML's SQL-based modeling with a full deployment platform that meets VPC and SLA requirements.

Practice this question →

50

Multi-Selectmedium

A research team wants to train a very large transformer model using Google's custom AI accelerators. They need the highest compute density and tightest interconnection for distributed training. Which THREE are true about Google's TPU infrastructure? (Choose 3)

Select 3 answers

A.TPU pods can scale to thousands of chips with low-latency interconnects

B.TPU v3 is the latest generation offering the best performance

C.TPU v4 pods offer high-bandwidth interconnects for large-scale distributed training

D.GPUs are Google's primary accelerator for large transformer training

E.TPU v5e is designed for cost-efficient inference and small-scale training

AnswersA, C, E

TPU pods are designed to scale to thousands of TPU chips with high-speed interconnects for efficient distributed training.

Why this answer

Option A is correct because TPU pods are designed to scale to thousands of TPU chips using a custom high-speed interconnect (e.g., the 2D torus mesh in TPU v2/v3 and the 3D torus in TPU v4), which provides low-latency, high-bandwidth communication essential for distributed training of very large transformer models. This architecture allows the research team to achieve the highest compute density and tightest interconnection for their workload.

Exam trap

The trap here is that candidates may assume TPU v3 is the latest generation because it is widely documented in older materials, but Google has since released v4, v5e, and v5p, with v5p being the current top performer for training.

Practice this question →

51

MCQmedium

A company wants to build a customer service chatbot that answers questions about their internal policy documents. The documents are updated monthly, and the team cannot afford to retrain a model each time. Which approach is MOST appropriate?

A.Use a larger foundation model with a longer context window and paste all documents into each prompt

B.Fine-tune a base LLM on the policy documents monthly

C.Train a custom model from scratch on the policy documents each month

D.Use Retrieval-Augmented Generation (RAG) with the policy documents indexed in a vector store

AnswerD

RAG retrieves relevant document chunks at query time, ensuring the chatbot always answers from the latest uploaded documents without any model retraining.

Why this answer

RAG (Retrieval-Augmented Generation) allows the LLM to retrieve relevant document sections at inference time, so knowledge stays current without retraining. The other options either require expensive retraining for each update or lack document grounding.

Practice this question →

52

MCQmedium

A healthcare startup needs to process medical claims and extract structured data (e.g., patient name, procedure codes, amounts) from scanned PDF forms. They require HIPAA compliance and prefer a pre-built Google Cloud service. Which service should they use?

A.Cloud Vision API with OCR

B.Vertex AI AutoML Tables

C.Document AI Healthcare API with the Claims processor

D.Natural Language AI with entity extraction

AnswerC

Document AI is HIPAA-eligible and offers a pre-trained Claims processor for medical claim forms.

Why this answer

Option C is correct because the Document AI Healthcare API with the Claims processor is a pre-built, HIPAA-compliant Google Cloud service specifically designed to extract structured data (e.g., patient name, procedure codes, amounts) from scanned medical claim forms. It leverages specialized machine learning models trained on healthcare documents, ensuring accurate parsing of fields like CPT codes and ICD-10 codes while meeting regulatory requirements.

Exam trap

The trap here is that candidates often confuse general OCR services (like Cloud Vision API) with domain-specific document processors, overlooking that HIPAA compliance and pre-built healthcare form parsing require a specialized service like Document AI Healthcare API rather than a generic text extraction tool.

How to eliminate wrong answers

Option A is wrong because Cloud Vision API with OCR only extracts raw text from images without understanding the document structure or healthcare-specific fields, and it lacks built-in HIPAA compliance for protected health information (PHI). Option B is wrong because Vertex AI AutoML Tables is for tabular data prediction (e.g., regression or classification on structured datasets), not for extracting structured data from scanned PDF forms. Option D is wrong because Natural Language AI with entity extraction is designed for analyzing unstructured text (e.g., clinical notes) and cannot process scanned PDF forms or extract structured fields like procedure codes from document layouts.

Practice this question →

53

MCQeasy

A data science team wants to run a machine learning model directly on data stored in BigQuery without moving the data to a separate environment. Which Google Cloud service should they use?

A.Vertex AI Training

B.Cloud Dataproc

C.Google AI Studio

D.BigQuery ML

AnswerD

BigQuery ML allows users to build and run ML models using SQL directly on BigQuery data, no data export needed.

Why this answer

BigQuery ML enables creating and executing ML models using standard SQL queries directly on data in BigQuery, eliminating data movement.

Practice this question →

54

MCQhard

An enterprise customer needs to ensure that all data sent to the Gemini API is not used by Google for model improvement and must support a HIPAA BAA. Which access tier should they use?

A.Google AI Studio (free tier)

B.Vertex AI (enterprise tier)

C.Gemini API via Google Workspace

D.Gemini API via API key (developer tier)

AnswerB

Enterprise tier provides data isolation, no training on customer data, and HIPAA BAA.

Why this answer

Vertex AI for enterprise offers data isolation, no use of customer data for training, and HIPAA BAA. Google AI Studio and Gemini API (developer) do not provide these guarantees.

Practice this question →

55

Multi-Selectmedium

A global e-commerce company wants to build a multilingual customer support chatbot that can understand queries in 50 languages and respond in the same language. They need to process both text and images (e.g., product photos) in the queries. Which THREE Google Cloud services should they consider? (Select 3 options.)

Select 2 answers

A.Translation API

B.Vision API

C.Natural Language API

D.Text-to-Speech

E.Gemini API (multimodal)

AnswersA, E

Translation API can translate between 100+ languages, supplementing Gemini's language support for less common languages.

Why this answer

Translation API is correct because it provides real-time language translation across 100+ languages, enabling the chatbot to understand and respond in 50 languages. It integrates directly with other Google Cloud services to maintain language consistency in a multilingual pipeline.

Exam trap

Cisco often tests the distinction between single-purpose APIs (like Vision API or Natural Language API) and multimodal models (like Gemini API) that can handle both text and images natively, leading candidates to over-select specialized services instead of the integrated multimodal solution.

Practice this question →

56

MCQhard

A company wants to run a large-scale training job for a 175B parameter model. They need to minimize training time and cost. Which TPU version and configuration should they choose?

A.TPU v2-8

B.TPU v3-32

C.TPU v4-64

D.TPU v5e-256

AnswerD

v5e provides a good balance of performance and cost for large-scale training.

Why this answer

Option D is correct because the TPU v5e-256 offers the best performance-per-dollar for large-scale training of a 175B parameter model. With 256 chips in a pod, it provides massive parallelism and high memory bandwidth, significantly reducing training time compared to earlier generations while maintaining cost efficiency through optimized architecture.

Exam trap

The trap here is that candidates often assume larger chip count alone (like v4-64) is sufficient, but fail to consider the memory capacity per chip and the cost-efficiency of newer generations, leading them to overlook the v5e-256's superior balance of scale and affordability.

How to eliminate wrong answers

Option A is wrong because TPU v2-8 provides only 8 chips with 64 GB HBM total, which is far too small for a 175B parameter model (requiring ~350 GB just for parameters) and lacks the memory capacity and inter-chip interconnect speed needed for efficient distributed training. Option B is wrong because TPU v3-32 offers 32 chips with 128 GB HBM total, still insufficient memory for a 175B model and uses older interconnects that create bottlenecks at scale, leading to longer training times and higher overall cost. Option C is wrong because TPU v4-64 provides 64 chips with 256 GB HBM total, which is still below the memory requirement for a 175B model and, while faster than v3, does not match the cost-efficiency and throughput of v5e at the required scale.

Practice this question →

57

MCQmedium

A healthcare startup needs to process sensitive patient data using NLP models on Google Cloud. They require HIPAA compliance and the ability to run models within their VPC. Which service should they use to access Gemini models?

A.Gemini API directly via API key

B.BigQuery ML

C.Vertex AI

D.Google AI Studio

AnswerC

Vertex AI offers VPC Service Controls, data isolation, audit logging, and can sign a HIPAA BAA.

Why this answer

Vertex AI provides enterprise-grade features including VPC Service Controls, data isolation, and HIPAA BAA. Google AI Studio is free-tier prototyping only and does not offer these compliance or security controls.

Practice this question →

58

Multi-Selectmedium

A data scientist wants to use Google Cloud AI services to build a solution that transcribes customer support calls, analyzes sentiment, and translates transcripts into multiple languages. Which TWO services are needed?

Select 2 answers

A.Vision AI

B.Natural Language AI

C.Cloud Translation API

D.Document AI

E.Cloud Speech-to-Text API

AnswersC, E

Required for translating transcripts.

Why this answer

Speech-to-Text transcribes audio, Translation API translates text. Natural Language AI could be used for sentiment but is not required for translation. Document AI and Vision AI are irrelevant.

Practice this question →

59

MCQmedium

A developer wants to add real-time speech transcription to a customer call center application. They need low latency and high accuracy for multiple languages. Which Google AI API is most appropriate?

A.Speech-to-Text API

B.Natural Language API

C.Text-to-Speech API

D.Translation API

AnswerA

Speech-to-Text API converts audio to text, supporting real-time streaming and multiple languages.

Why this answer

The Speech-to-Text API is the correct choice because it is specifically designed to convert audio into text in real time, supporting over 125 languages and variants with low-latency streaming. It offers features like automatic punctuation, speaker diarization, and domain-specific models (e.g., phone call) that directly meet the requirements of a customer call center application needing high accuracy across multiple languages.

Exam trap

Cisco often tests the distinction between APIs that process text (Natural Language, Translation) versus those that process audio (Speech-to-Text, Text-to-Speech), and the trap here is confusing the direction of conversion (speech-to-text vs. text-to-speech) or assuming a translation API can handle raw audio input.

How to eliminate wrong answers

Option B is wrong because the Natural Language API analyzes text for entities, sentiment, and syntax, but it does not process audio or perform speech transcription. Option C is wrong because the Text-to-Speech API converts text into spoken audio, which is the opposite direction of the required speech-to-text functionality. Option D is wrong because the Translation API translates text between languages but cannot transcribe speech from audio input.

Practice this question →

60

Multi-Selectmedium

A machine learning team is using Vertex AI to train a custom model. They want to optimize hyperparameters automatically. Which TWO steps are necessary to set up hyperparameter tuning in Vertex AI? (Choose TWO)

Select 2 answers

A.Enable Vertex AI Experiments

B.Use a custom container with a GPU

C.Enable distributed training across multiple nodes

D.Define a hyperparameter metric in the training code

E.Create a HyperparameterTuningJob with parameter specifications

AnswersD, E

The training code must report a metric that Vertex AI uses to evaluate hyperparameter trials.

Why this answer

To run hyperparameter tuning, you must specify a hyperparameter metric in the training code and configure the tuning job in Vertex AI with the parameter specifications.

Practice this question →

61

MCQeasy

A data analyst needs to run a simple regression model directly on data stored in BigQuery without moving data to another platform. Which service should they use?

A.TensorFlow on Compute Engine

B.BigQuery ML

C.Vertex AI Training

D.Google Colab

AnswerB

BigQuery ML enables ML via SQL on BigQuery data.

Why this answer

BigQuery ML (B) is correct because it allows users to create and execute machine learning models using standard SQL syntax directly on data stored in BigQuery, without needing to export data to a separate platform. This service is specifically designed for running regression, classification, and other models natively within BigQuery, leveraging its serverless architecture and built-in ML capabilities.

Exam trap

The trap here is that candidates often confuse Vertex AI Training (a full-featured ML platform) with BigQuery ML, not realizing that Vertex AI requires data export and more setup, while BigQuery ML is purpose-built for in-database modeling with minimal overhead.

How to eliminate wrong answers

Option A is wrong because TensorFlow on Compute Engine requires moving data out of BigQuery to a virtual machine, where you must manually manage infrastructure, install dependencies, and write custom training code, which contradicts the requirement of not moving data. Option C is wrong because Vertex AI Training is a managed ML platform that typically requires exporting data from BigQuery to Cloud Storage or a dataset in Vertex AI, and it involves more complex pipeline setup than a simple regression model. Option D is wrong because Google Colab is a Jupyter notebook environment that runs in the cloud but requires data to be loaded from BigQuery into a DataFrame, moving it out of BigQuery's native storage, and it does not provide a direct SQL-based modeling interface.

Practice this question →

62

MCQmedium

A data scientist wants to run large-scale distributed training of a custom deep learning model using Google's custom AI accelerators. Which infrastructure should they choose to minimize cost while leveraging Google's proprietary chips?

A.Cloud TPU v5e

B.Compute Engine with NVIDIA A100 GPUs

C.Vertex AI Workbench with custom machines

D.Google Colab Pro with TPU runtime

AnswerA

Cloud TPU v5e is Google's custom AI accelerator, optimized for large-scale training and cost-efficient for many models.

Why this answer

Cloud TPU v5e is the correct choice because it is Google's proprietary custom AI accelerator designed specifically for large-scale distributed training of deep learning models, offering superior cost-efficiency compared to GPUs for many workloads. TPU v5e provides a balanced price-performance ratio for medium-to-large training tasks, and Google's TPU architecture is optimized for TensorFlow and JAX, enabling efficient scaling across multiple TPU pods. This minimizes cost while leveraging Google's custom chips, as opposed to using NVIDIA GPUs which are not Google's proprietary hardware.

Exam trap

The trap here is that candidates may confuse 'custom AI accelerators' with any high-performance hardware like GPUs, but the question specifically requires Google's proprietary chips (TPUs), and Cloud TPU v5e is the only option that directly provides cost-optimized, large-scale distributed training using Google's own accelerators.

How to eliminate wrong answers

Option B is wrong because Compute Engine with NVIDIA A100 GPUs uses third-party hardware (NVIDIA) rather than Google's proprietary chips, and while powerful, it typically incurs higher costs for large-scale distributed training compared to TPUs for suitable workloads. Option C is wrong because Vertex AI Workbench with custom machines is a development environment for building and training models, not a specific infrastructure choice for leveraging Google's custom AI accelerators; it can use TPUs or GPUs but does not inherently minimize cost with proprietary chips. Option D is wrong because Google Colab Pro with TPU runtime is designed for small-scale experimentation and prototyping, not for large-scale distributed training, and it lacks the scalability and cost efficiency of Cloud TPU v5e for production workloads.

Practice this question →

63

MCQmedium

A data analyst wants to build a regression model in BigQuery to predict sales from historical data without writing any Python code. Which BigQuery ML statement should they use to define the model?

A.CREATE MODEL my_model OPTIONS(model_type='linear_reg') AS SELECT ...

B.INSERT INTO model my_model VALUES ...

C.CREATE ML my_model AS (SELECT ...)

D.SELECT ML.TRAIN('linear_reg', ...)

AnswerA

This is the correct syntax to create a linear regression model in BigQuery ML.

Why this answer

The CREATE MODEL statement with option model_type='linear_reg' is used to create a linear regression model in BigQuery ML.

Practice this question →

64

MCQmedium

A team is developing a mobile app that must run AI inference on-device for low latency and offline capability. Which Gemini model variant is designed specifically for on-device deployment?

A.Gemini Pro

B.Gemini Nano

C.Gemini Ultra

D.Gemini Flash

AnswerB

Gemini Nano is specifically designed for on-device inference, offering efficiency with low latency and offline capability.

Why this answer

Gemini Nano is the smallest and most efficient model in the Gemini family, specifically optimized for on-device deployment. It is designed to run directly on mobile devices (e.g., Android phones) using hardware acceleration like Google's Pixel Neural Core or Qualcomm's AI Engine, enabling low-latency inference and offline capability without requiring a cloud connection.

Exam trap

The trap here is that candidates confuse 'lightweight cloud model' (Gemini Flash) with 'on-device model' (Gemini Nano), assuming any 'fast' or 'small' variant is suitable for mobile deployment, but Flash still requires cloud connectivity and is not optimized for local hardware constraints.

How to eliminate wrong answers

Option A is wrong because Gemini Pro is a mid-size model intended for cloud-based, high-performance tasks such as complex reasoning and multimodal analysis, not for on-device deployment due to its larger memory and compute requirements. Option C is wrong because Gemini Ultra is the largest and most capable model, designed for enterprise-scale cloud workloads and advanced research, making it unsuitable for resource-constrained mobile devices. Option D is wrong because Gemini Flash is a lightweight cloud model optimized for speed and cost in cloud inference, but it is not purpose-built for on-device execution and still requires a network connection.

Practice this question →

65

MCQmedium

An enterprise needs to generate natural-sounding speech from text for a voice assistant. They require low latency and support for custom voice models. Which service should they use?

A.Text-to-Speech API

B.Cloud Translation API

C.Vertex AI Text Generation

D.Speech-to-Text API

AnswerA

Text-to-Speech API converts text into natural-sounding speech, supports low latency and custom voice models.

Why this answer

The Text-to-Speech API (A) is correct because it is specifically designed to convert text into natural-sounding speech with low latency, and it supports custom voice models through features like Custom Voice and WaveNet voices. This directly meets the enterprise's requirements for a voice assistant that needs real-time, high-quality speech synthesis.

Exam trap

The trap here is confusing the Text-to-Speech API with the Speech-to-Text API, as candidates often mix up the direction of conversion (text-to-audio vs. audio-to-text) under time pressure.

How to eliminate wrong answers

Option B (Cloud Translation API) is wrong because it translates text between languages, not text to speech, and does not generate audio output. Option C (Vertex AI Text Generation) is wrong because it generates text content (e.g., chat responses, summaries) rather than synthesizing speech from text. Option D (Speech-to-Text API) is wrong because it performs the inverse operation—converting audio speech into text—and does not produce speech output.

Practice this question →

66

MCQhard

A company is evaluating SLA guarantees for a generative AI model deployed on Vertex AI. They require at least 99.9% uptime for production inference. Which SLA tier should they select?

A.Serverless endpoint with min replicas=0

B.Batch prediction job

C.Regional endpoint with at least one node

D.Global endpoint with automatic scaling

AnswerC

Regional endpoints with at least one node qualify for 99.9% SLA for online prediction.

Why this answer

Regional endpoints with at least one node guarantee a minimum of 99.9% uptime for production inference because they maintain a dedicated, always-on compute resource. This SLA applies to regional endpoints with a minimum of one replica, ensuring the model is continuously available for serving requests without cold-start delays.

Exam trap

The trap here is that candidates confuse automatic scaling or serverless endpoints with high availability, not realizing that only regional endpoints with a minimum of one replica meet the 99.9% SLA requirement for production inference.

How to eliminate wrong answers

Option A is wrong because a serverless endpoint with min replicas=0 can scale down to zero, introducing cold-start latency and not guaranteeing 99.9% uptime as the endpoint may be unavailable during scale-up. Option B is wrong because batch prediction jobs are not designed for real-time inference and do not offer an uptime SLA; they are asynchronous and may fail or queue without availability guarantees. Option D is wrong because global endpoints with automatic scaling do not have a specific SLA tier for 99.9% uptime; they are optimized for latency and load distribution but lack the dedicated replica requirement needed for the highest SLA commitment.

Practice this question →

67

Multi-Selectmedium

An enterprise wants to use Gemini for a customer-facing application. They require the following: data isolation in a VPC, audit logging, and SLA guarantees. Which THREE features of Vertex AI satisfy these requirements?

Select 3 answers

A.Vertex AI SLA (Service Level Agreement)

B.Gemini Nano on-device

C.VPC Service Controls

D.Cloud Audit Logs integration

E.Google AI Studio free tier

AnswersA, C, D

SLA guarantees uptime and performance.

Why this answer

Option A is correct because Vertex AI offers a defined Service Level Agreement (SLA) that guarantees uptime and performance metrics for enterprise customers, which is a core requirement for customer-facing applications. The SLA provides contractual assurances, typically covering availability and response times, ensuring the enterprise can meet its own service commitments.

Exam trap

Cisco often tests the distinction between development tools (like AI Studio free tier) and production-ready enterprise features, and candidates mistakenly assume that any Google AI offering includes SLA and VPC controls by default.

Practice this question →

68

MCQmedium

A financial services firm needs to fine-tune a large language model on proprietary financial data. They require data to never leave their VPC and need full audit logging. Which Gemini access method should they use?

A.Vertex AI

B.Google AI Studio

C.Gemini API directly via Cloud Endpoints

D.Model Garden in Colab

AnswerA

Vertex AI offers VPC Service Controls, audit logging, and data isolation.

Why this answer

Vertex AI is the correct access method because it is the only option that allows fine-tuning of Gemini models within a customer's VPC (Virtual Private Cloud) with full audit logging via Cloud Audit Logs. This ensures proprietary financial data never leaves the secure network boundary, meeting strict compliance and data residency requirements.

Exam trap

The trap here is that candidates may confuse Google AI Studio's free-tier accessibility with enterprise-grade security, overlooking that Vertex AI is the only option with VPC controls and audit logging for fine-tuning proprietary data.

How to eliminate wrong answers

Option B is wrong because Google AI Studio is a web-based prototyping tool that does not support VPC-scoped fine-tuning or enterprise-grade audit logging; data is processed on Google's infrastructure outside the customer's VPC. Option C is wrong because the Gemini API directly via Cloud Endpoints does not provide native VPC controls or fine-tuning capabilities; it is a stateless API call without persistent model customization. Option D is wrong because Model Garden in Colab is a discovery and experimentation environment that lacks VPC isolation, audit logging, and fine-tuning support for production workloads.

Practice this question →

69

MCQeasy

Which Google Cloud AI service is specifically designed for extracting structured data from scanned documents, such as invoices and receipts?

A.Document AI

B.Natural Language AI

C.Translation AI

D.Vision AI

AnswerA

Document AI is purpose-built for extracting structured data from scanned documents using OCR and parsing.

Why this answer

Document AI is the correct answer because it is purpose-built for understanding and extracting structured data from unstructured documents like invoices, receipts, and forms. It uses specialized processors (e.g., the Invoice Parser or Expense Parser) that combine optical character recognition (OCR) with natural language understanding and machine learning models trained on document layouts, enabling it to output structured fields such as vendor name, total amount, and line items.

Exam trap

The trap here is that candidates often confuse Vision AI’s general OCR capability with Document AI’s specialized document understanding, overlooking that Vision AI cannot natively extract structured fields like line items or totals without extensive custom coding.

How to eliminate wrong answers

Option B is wrong because Natural Language AI is designed for analyzing and extracting insights from text (e.g., sentiment, entity recognition, syntax analysis), not for processing scanned document images or extracting structured data from forms. Option C is wrong because Translation AI is a neural machine translation service that converts text between languages, with no capability to parse scanned documents or extract structured fields. Option D is wrong because Vision AI provides general-purpose image analysis (e.g., object detection, OCR for text extraction), but it lacks the specialized document understanding and pre-trained models for extracting structured data from invoices and receipts that Document AI offers.

Practice this question →

70

MCQhard

An enterprise is comparing Google Cloud Vertex AI vs AWS Bedrock vs Azure OpenAI for a generative AI application. Which unique Google differentiator allows the model to reference up-to-date web information and private data with managed retrieval?

A.Vertex AI Agent Builder with search grounding

B.TPU availability

C.Integration with Google Workspace

D.Multimodal understanding

AnswerA

Agent Builder provides managed search grounding.

Why this answer

Vertex AI offers grounding with Google Search and private data sources, a capability not directly matched by AWS Bedrock or Azure OpenAI.

Practice this question →

71

MCQmedium

A company wants to build a customer service chatbot that answers questions about their internal policy documents. The documents are updated monthly, and the team cannot afford to retrain a model each time. Which approach is MOST appropriate?

A.Fine-tune a base LLM on the policy documents monthly

B.Use a larger foundation model with a longer context window and paste all documents into each prompt

C.Train a custom model from scratch on the policy documents each month

D.Use Retrieval-Augmented Generation (RAG) with the policy documents indexed in a vector store

AnswerD

RAG retrieves relevant document chunks at query time, ensuring the chatbot always answers from the latest uploaded documents without any model retraining.

Why this answer

Retrieval-Augmented Generation (RAG) is the most appropriate approach because it allows the chatbot to answer questions based on the latest policy documents without retraining the model. By indexing the documents in a vector store and retrieving relevant chunks at query time, RAG ensures the model uses up-to-date information while keeping the underlying LLM static, which is cost-effective and scalable for monthly updates.

Exam trap

Cisco often tests the misconception that fine-tuning is the only way to incorporate new information, but the trap here is that candidates overlook RAG's ability to handle dynamic data without retraining, confusing 'model adaptation' with 'data retrieval'.

How to eliminate wrong answers

Option A is wrong because fine-tuning a base LLM monthly on new policy documents is expensive, time-consuming, and risks catastrophic forgetting, where the model loses previously learned information. Option B is wrong because pasting all documents into each prompt exceeds the context window limits of even the largest models (e.g., 128K tokens), leading to truncation, high latency, and increased cost per query. Option C is wrong because training a custom model from scratch each month is prohibitively expensive, requires massive computational resources and data, and is unnecessary when a pre-trained LLM with RAG can achieve the same goal.

Practice this question →

72

MCQhard

A financial institution needs to deploy a large language model (LLM) for a customer-facing application that must comply with HIPAA and have strict data residency controls. They also require the ability to ground responses in real-time search results from the web. Which combination of services should they use?

A.Google AI Studio with Gemini Pro

B.Vertex AI with Gemini Pro and Grounding with Google Search

C.Gemini API directly from AI Studio with custom VPC

D.Vertex AI with Gemini and Amazon Bedrock for grounding

AnswerB

Vertex AI offers enterprise-grade security, compliance (HIPAA BAA), and supports grounding with Google Search for real-time information retrieval.

Why this answer

Vertex AI with Gemini Pro and Grounding with Google Search is correct because it provides a HIPAA-compliant platform (Vertex AI) with the ability to ground LLM responses in real-time web search results via Google Search Grounding, while also supporting data residency controls through customer-managed encryption keys and regional endpoints. This combination meets all requirements: compliance, data residency, and real-time grounding.

Exam trap

Cisco often tests the distinction between prototyping tools (AI Studio) and production platforms (Vertex AI), and the trap here is assuming that any Google AI service can be used for HIPAA-compliant deployment without checking for enterprise features like VPC Service Controls and data residency support.

How to eliminate wrong answers

Option A is wrong because Google AI Studio is a prototyping tool, not a production deployment platform, and does not support HIPAA compliance or data residency controls. Option C is wrong because the Gemini API directly from AI Studio lacks enterprise features like VPC Service Controls for data residency and does not offer HIPAA-compliant deployment; custom VPC alone does not satisfy HIPAA requirements. Option D is wrong because Amazon Bedrock is an AWS service, not part of the Google AI ecosystem, and mixing Vertex AI with Bedrock introduces cross-cloud complexity that violates the single-vendor strategy implied by the exam domain and does not provide native Google Search Grounding.

Practice this question →

73

MCQeasy

Which Google AI research organization is responsible for AlphaFold, a breakthrough in protein structure prediction?

A.Google Research

B.Google DeepMind

C.Google Brain

D.X Development

AnswerB

DeepMind developed AlphaFold.

Why this answer

Google DeepMind is the research lab behind AlphaFold, AlphaCode, and other notable achievements.

Practice this question →

74

MCQeasy

A startup wants to quickly prototype a conversational AI application using Gemini. They need free access during development and do not require VPC controls. Which access tier should they choose?

A.Vertex AI on a pay-as-you-go basis

B.Vertex AI with VPC-SC

C.Gemini API via Cloud Run

D.Google AI Studio free tier

AnswerD

Google AI Studio provides free prototyping without VPC controls.

Why this answer

Option D is correct because Google AI Studio's free tier provides free access to Gemini models for rapid prototyping without requiring VPC controls or any billing setup. This aligns directly with the startup's need for quick, cost-free development iteration before moving to production.

Exam trap

Cisco often tests the misconception that any Google Cloud service requires a billing account, but Google AI Studio's free tier explicitly bypasses this for prototyping, while options like Vertex AI or Cloud Run always incur costs even at low usage.

How to eliminate wrong answers

Option A is wrong because Vertex AI on a pay-as-you-go basis incurs costs from the start, which contradicts the requirement for free access during development. Option B is wrong because Vertex AI with VPC-SC adds unnecessary VPC security controls and costs, which the startup explicitly does not need. Option C is wrong because the Gemini API via Cloud Run requires a billing account and incurs compute and API usage costs, making it not free for prototyping.

Practice this question →

75

MCQmedium

A company wants to build a customer service chatbot that answers questions about their internal policy documents. The documents are updated monthly, and the team cannot afford to retrain a model each time. Which approach is MOST appropriate?

A.Fine-tune a base LLM on the policy documents monthly

B.Use a larger foundation model with a longer context window and paste all documents into each prompt

C.Train a custom model from scratch on the policy documents each month

D.Use Retrieval-Augmented Generation (RAG) with the policy documents indexed in a vector store

AnswerD

RAG retrieves relevant document chunks at query time, ensuring the chatbot always answers from the latest uploaded documents without any model retraining.

Why this answer

Retrieval-Augmented Generation (RAG) is the most appropriate approach because it allows the chatbot to answer questions using the latest policy documents without retraining the model. By indexing the documents in a vector store and retrieving relevant chunks at query time, RAG ensures the model's responses are grounded in the most current information, even as documents are updated monthly. This avoids the cost and complexity of fine-tuning or retraining a model each time the documents change.

Exam trap

Cisco often tests the misconception that fine-tuning or retraining is necessary for domain-specific knowledge, when in fact RAG provides a cost-effective, dynamic alternative that avoids model updates and maintains up-to-date information.

How to eliminate wrong answers

Option A is wrong because fine-tuning a base LLM on the policy documents monthly would require significant computational resources and time for each update, and the model may still produce outdated or hallucinated answers if the fine-tuning data is not perfectly aligned with the latest documents. Option B is wrong because pasting all policy documents into each prompt would exceed the context window limits of even the largest foundation models (e.g., 128K tokens for GPT-4 Turbo), leading to truncated inputs, high latency, and increased cost per query. Option C is wrong because training a custom model from scratch each month is prohibitively expensive, time-consuming, and requires large amounts of training data and expertise, making it impractical for monthly document updates.

Practice this question →

Page 1 of 2 · 100 questions totalNext →

Ready to test yourself?

Try a timed practice session using only Gcl Google Ai Ecosystem questions.

Start 20-question session