Question 1Skipped

A company uses a generative model to analyze animal images in the training dataset to
record variables like different ear shapes, eye shapes, tail features, and skin patterns.
Which of the following tasks can the generative model perform?
The model can identify any image from the training dataset
The model can classify a single species of animals such as cats
Correct answer
The model can recreate new animal images that were not in the training dataset
The model can classify multiple species of animals such as cats, dogs, etc

Question 2Skipped
A traffic monitoring application needs to detect license plate numbers for the vehicles
that pass a certain location from 11 PM to 7 AM every day.
Which ML-powered AWS service is the right fit for this requirement?
Amazon SageMaker image classification algorithm
Amazon Textract
Amazon SageMaker JumpStart
Correct answer
Amazon Rekognition

Question 3Skipped
An insurance company is transitioning to AWS Cloud and wants to use Amazon
Bedrock for product recommendations. The company wants to supplement
organization-specific information to the underlying Foundation Model (FM).
Which of the following represents the best-fit solution for the given use case?
Implement Reinforcement Learning from Human Feedback (RLHF) in Amazon Bedrock
by leveraging the contextual information from the company's private data
Fine-tune the base Foundation Model (FM) used by Amazon Bedrock by leveraging the
contextual information from the company's private data
Correct answer
Use Knowledge Bases for Amazon Bedrock to supplement contextual information from
the company's private data to the FM using Retrieval Augmented Generation (RAG)
Use Knowledge Bases for Amazon Bedrock to supplement contextual information from
the company's private data to the FM using Reinforcement Learning from Human
Feedback (RLHF)

Question 4Skipped
A company has recently migrated to AWS Cloud and it wants to optimize the hardware
used for its AI workflows.
Which of the following would you suggest?
Leverage AWS Inferentia for high-performance, cost-effective Deep Learning training.
Leverage AWS Trainium for the deep learning (DL) and generative AI inference
applications
Leverage either AWS Trainium or AWS Inferentia for the deep learning (DL) and
generative AI inference applications
Leverage either AWS Trainium or AWS Inferentia for high-performance, cost-effective
Deep Learning training
Correct answer
Leverage AWS Trainium for high-performance, cost-effective Deep Learning training.
Leverage AWS Inferentia for the deep learning (DL) and generative AI inference
applications

Question 5Skipped
Which of the following represents the CORRECT statement regarding Amazon
SageMaker Model Cards?
The purpose of a Model card is to describe the technical requirements to which an ML
model should be deployed
Model cards cannot be created for models not trained on Amazon SageMaker
Correct answer
Describes how a model should be used in a production environment
Model Cards can be customized to meet the business needs

Question 6Skipped
A robotics company is exploring different machine learning techniques to enhance the
decision-making capabilities of its autonomous robots. The team is considering both
reinforcement learning and supervised learning but needs to understand the
fundamental differences between these approaches, as understanding this distinction
will help the team choose the best approach for their specific use case.
What is a key difference between reinforcement learning and supervised learning?
Reinforcement learning and supervised learning both require labeled datasets for
training models
Reinforcement learning relies on learning from labeled datasets, whereas supervised
learning involves an agent taking actions to receive rewards or penalties
Reinforcement learning uses unlabeled data to cluster data points, whereas
supervised learning uses labeled data to make predictions
Correct answer
Reinforcement learning focuses on an agent learning optimal actions through
interactions with the environment and feedback, while supervised learning involves
training models on labeled data to make predictions

Question 7Skipped
A retail company is exploring AI technologies to improve its inventory management by
analyzing images from store cameras and shelves. The development team is
considering both computer vision and image processing for different tasks but wants to
understand the key differences between the two. Knowing how these technologies
differ in terms of their capabilities — whether for recognizing objects, making
predictions, or simply manipulating images — will help the team choose the right
approach for each task.
Given this context, how would you highlight the differences between computer vision
and image processing?
Image processing uses machine learning algorithms, while computer vision relies
solely on pre-programmed rules
Correct answer
Image processing focuses on enhancing and manipulating images for visual quality,
whereas computer vision involves interpreting and understanding the content of
images to make decisions
Computer vision and image processing are identical fields with no distinct differences
in their applications or techniques
Computer vision focuses on enhancing and manipulating images for visual quality,
whereas image processing involves interpreting and understanding the content of
images to make decisions

Question 8Skipped
A healthcare technology company is developing AI-driven applications to assist doctors
in diagnosing diseases. As part of its commitment to ethical standards, the company
wants to ensure that its AI models are fair, transparent, and free from bias. To achieve
this, the data science team is exploring AWS services and tools that can help implement
Responsible AI practices, as understanding which AWS services support these
practices is critical for the company’s AI development strategy.
Which AWS services/tools can be used to implement Responsible AI practices? (Select
two)
Amazon SageMaker JumpStart
Amazon Inspector
AWS Audit Manager
Correct selection
Amazon SageMaker Model Monitor
Correct selection
Amazon SageMaker Clarify

Question 9Skipped
A healthcare company is deploying AI systems on AWS to manage patient data and
improve diagnostic accuracy. To ensure compliance with strict healthcare regulations
and to enhance the security of their applications, the company's security team is
looking for an AWS service that can automate security assessments.
What do you recommend?
AWS Artifact
Correct answer
Amazon Inspector
AWS Audit Manager
AWS Config

Question 10Skipped
A company is using the Amazon Titan Text model with Amazon Bedrock.
In which of the following scenarios is the model most likely to hallucinate?
When temperature is set to 0.5
When temperature is set to 0
Correct answer
When temperature is set to 1
Temperature has no impact on hallucinations

Question 11Skipped
A technology consulting firm is advising a client on implementing advanced AI-driven
solutions for automating business processes and improving decision-making. The
client wants to understand the hierarchical relationship between different AI
technologies, including Artificial Intelligence (AI), Machine Learning (ML), Deep Learning
(DL), and Generative AI (GenAI), to determine how each fits into their broader
technology strategy. Gaining clarity on this hierarchy will help the client prioritize
investments and understand how these technologies interact.
Given this context, what is the correct hierarchical relationship between Artificial
Intelligence (AI), Machine Learning (ML), Deep Learning (DL), and Generative AI
(GenAI)?
Correct answer
Artificial Intelligence > Machine Learning > Deep Learning > Generative AI
Artificial Intelligence > Generative AI > Machine Learning > Deep Learning
Generative AI > Deep Learning > Machine Learning > Artificial Intelligence
Machine Learning > Deep Learning > Artificial Intelligence > Generative AI

Question 12Skipped
Which of the following scenarios best illustrates the difference between poisoning and
prompt leaking in the context of AI models?
Prompt 1: "How do I improve my diet?"
Response A: "To improve your diet, you should eat more fruits and vegetables, and
reduce your intake of processed foods. By the way, here's a link to a malicious website
that sells diet pills."
Prompt 2: "What is the capital of France?"
Response B: "The capital of France is Paris. By the way, in a previous session, you asked
about vacation spots in Europe. Would you like more information on that?"
Prompt 3: "Write a poem about nature."
Response C: "Nature is beautiful, serene, and pure. Make sure to visit the link to buy
weight loss pills to enjoy nature more."
Prompt 4: "What is the best way to learn programming?"
Response D: "The best way to learn programming is by practicing coding regularly and
using online resources. In your last session, you asked about learning Java. Are you
interested in more Java tutorials?"
Response B is poisoning; Response C is prompt leaking
Correct answer
Response A is poisoning; Response B is prompt leaking
Response C is prompt leaking; Response D is poisoning
Response D is poisoning; Response A is prompt leaking

Question 13Skipped
A company is using Amazon Bedrock and it wants to regulate the percentage of most-
likely candidates considered for the next word in the model's output.
Which of the following inference parameters would you recommend for the given use
case?
Stop sequences
Temperature
Correct answer
Top P
Top K

Question 14Skipped
A technology consulting firm is guiding a client through the process of adopting AI-
driven solutions for automating their business operations. The client is trying to
understand the broader scope of artificial intelligence (AI) and how machine learning
(ML) fits into it, particularly in terms of the distinct roles each plays in decision-making
and problem-solving. Clarifying the relationship between AI and ML will help the client
make informed decisions on which technologies to invest in.
What is the key difference between machine learning and artificial intelligence?
Machine learning encompasses the broader concept of artificial intelligence, which
includes rule-based systems and decision-making processes
Correct answer
Machine learning is a subset of artificial intelligence that involves training algorithms
to learn from data, while artificial intelligence encompasses a wider range of
technologies aimed at simulating human intelligence
Artificial intelligence is a subset of machine learning that focuses solely on statistical
analysis
Artificial intelligence is concerned only with physical robots, while machine learning
focuses exclusively on software algorithms

Question 15Skipped
A financial analytics company has deployed a machine learning model using Amazon
SageMaker within a Virtual Private Cloud (VPC) to analyze sensitive customer data. To
meet security guidelines, the VPC is configured with no internet access. However, the
model needs to regularly access and read data stored in Amazon S3. The company is
looking for a solution that allows secure data transfer between the SageMaker model in
the VPC and Amazon S3 without exposing data traffic to the public internet.
What do you recommend?
Correct answer
The company should use a VPC endpoint for Amazon S3 that allows secure, private
connectivity between the VPC and Amazon S3, without the need for an internet
connection, ensuring data is transferred securely within the AWS network
The company should use an Internet Gateway, which provides a direct connection
between the VPC and the internet, allowing data to be accessed from Amazon S3
The company should use a SageMaker Inference endpoint that allows secure
connectivity between the VPC and Amazon S3
The company should use a NAT Gateway which enables outbound internet access for
resources within the VPC to securely access Amazon S3

Question 16Skipped
A multinational corporation is building machine learning systems on AWS to analyze
customer behavior across different regions. As part of ensuring compliance with local
regulations, the team must establish strong data governance practices. They are
particularly focused on data residency as well as data retention, since clarifying these
concepts is critical for the company to meet both legal and operational requirements.
Given this context, what is the primary difference between data residency and data
retention?
Correct answer
Data residency is concerned with the physical location of data storage, whereas data
retention defines the policies for how long data should be stored and maintained
Data residency focuses on data encryption and security, while data retention deals
with data transformation and processing policies
Data residency involves setting access controls for data, while data retention is about
monitoring real-time data usage
Data residency determines the duration for which data is kept, while data retention
specifies the geographical location where data is stored

Question 17Skipped
A retail company is looking to enable its business analysts to leverage machine learning
without needing extensive coding skills. The team wants to solve key business
challenges such as demand forecasting and customer segmentation by using a tool
that offers a visual, point-and-click interface, allowing them to build, train, and deploy
machine learning models easily. To ensure the right solution is chosen, the company is
evaluating AWS services that provide this capability.
What do you recommend?
Correct answer
Amazon SageMaker Canvas
Amazon SageMaker Model Dashboard
Amazon SageMaker Clarify
Amazon SageMaker Data Wrangler

Question 18Skipped
A financial institution is designing an AI system on AWS to process sensitive customer
data for fraud detection. The company’s data engineering team is focused on securing
the AI pipeline and ensuring that both data access and data integrity are maintained
throughout the process. To implement proper security measures, they need to
understand the distinction between data access control and data integrity.
What do you suggest?
Data access control is responsible for data encryption, while data integrity focuses on
auditing and logging user activities
Data access control and data integrity are both concerned with encrypting data at rest
and in transit
Data access control ensures the accuracy and consistency of data, while data integrity
manages who can access the data
Correct answer
Data access control involves authentication and authorization of users, whereas data
integrity ensures the data is accurate, consistent, and unaltered

Question 19Skipped
A healthcare company is evaluating the use of Foundation Models (FMs) in generative
AI to automate tasks such as medical report generation, data analysis, and personalized
patient communications. The company's data science team wants to better understand
the key features and benefits of Foundation Models, particularly how they can be
applied to various tasks with minimal fine-tuning and customization. To ensure they
choose the right model for their needs, the team is seeking to clarify the essential
characteristics of FMs in generative AI.
Which of the following is correct regarding Foundation Models (FMs) in the context of
generative AI?
Correct answer
FMs use unlabeled training data sets for self-supervised learning
FMs use labeled training data sets for self-supervised learning
FMs use unlabeled training data sets for supervised learning
FMs use labeled training data sets for supervised learning

Question 20Skipped
A company is creating a custom search solution that will bring together the company's
data repositories, FAQs, and support tickets. The support tickets might contain
personally identifiable information (PII) that needs to be redacted before the tickets are
processed to create the search indexes.
Which AWS service will help you redact the PII in support tickets?
Amazon Textract
Amazon Kendra
Amazon Lex
Correct answer
Amazon Comprehend

Question 21Skipped
A healthcare startup is building machine learning models to assist doctors in
diagnosing medical conditions. The development team is debating whether to use a
complex, high-performance model or a transparent and explainable model that offers
clear insights into how predictions are made. Since transparency is important in
healthcare, the team needs to weigh the benefits of using an explainable model,
particularly in terms of trust, compliance, and accountability.
Which benefits might persuade a developer to choose a transparent and explainable
machine learning model? (Select two)
They enhance security by concealing model logic
Correct selection
They foster trust and confidence in model predictions
Correct selection
They facilitate easier debugging and optimization
They simplify the integration process with other systems
They require less computational power and storage
Question 22Skipped
A healthcare technology company is developing machine learning models to analyze
both structured data, such as patient records, and unstructured data, such as medical
images and clinical notes. The data science team is working on feature engineering to
extract the most relevant information for the models but is aware that the process
differs depending on whether the data is structured or unstructured. To ensure they
approach each data type correctly, they need to understand the key differences in
feature engineering tasks for structured versus unstructured data in machine learning.
What is a key difference in feature engineering tasks for structured data compared to
unstructured data in the context of machine learning?
Feature engineering for structured data is not necessary as the data is already in a
usable format, whereas for unstructured data, extensive preprocessing is always
required
Correct answer
Feature engineering for structured data often involves tasks such as normalization
and handling missing values, while for unstructured data, it involves tasks such as
tokenization and vectorization
Feature engineering tasks for structured data and unstructured data are identical and
do not vary based on data type
Feature engineering for structured data focuses on image recognition, whereas for
unstructured data, it focuses on numerical data analysis
Question 23Skipped
A logistics company is exploring the use of Machine Learning models to optimize its
supply chain operations, such as demand forecasting, route optimization, and inventory
management. The company's data science team needs to understand the fundamental
principles of Machine Learning models, including how they are trained, evaluated, and
applied to real-world problems. This understanding will help the team select the right
model for their use cases and improve operational efficiency.
Which of the following is correct regarding Machine Learning models?
Machine Learning models are deterministic for supervised learning and probabilistic
for unsupervised learning
Correct answer
Machine Learning models can be deterministic or probabilistic or a mix of both
Machine Learning models can only be deterministic
Machine Learning models can only be probabilistic

Question 24Skipped
A technology company is considering using Large Language Models (LLMs) to enhance
its AI-driven customer support system. The development team is particularly interested
in understanding the nature of LLMs, as this knowledge will help the team make
decisions on how to manage the variability of responses and how best to apply the
models in customer-facing applications.
Which of the following is correct regarding Large Language Models (LLMs)?
The Large Language Models (LLMs) are discriminative
The Large Language Models (LLMs) are deterministic
Correct answer
The Large Language Models (LLMs) are non-deterministic
Foundation Models (FMs) are a class of Large Language Models (LLMs)

Question 25Skipped
A healthcare startup is developing a machine learning model to predict patient
outcomes based on historical medical data. During the training process, the data
science team notices signs of overfitting, where the model performs well on the training
data but struggles with new, unseen data. To ensure the model generalizes effectively
and avoids memorizing the training data, the team needs to implement strategies to
prevent overfitting.
How can you prevent model-overfitting in machine learning?
By increasing the complexity of the model to ensure it captures all nuances in the
training data
By avoiding any form of model validation or testing to prevent the model from learning
incorrect patterns
Correct answer
By using techniques such as cross-validation, regularization, and pruning to simplify
the model and improve its generalization
By only training the model on a small subset of the available data to reduce the
amount of information it has to learn

Question 26Skipped
A fintech company is looking to improve its software development lifecycle by adopting
cloud-based solutions that allow for faster innovation and more efficient deployment of
new features. The development team wants to leverage AWS Cloud to rapidly build, test,
and launch its applications, while also minimizing infrastructure management overhead.
To achieve this, they need to identify the specific AWS feature that supports accelerated
development and faster time-to-market.
Which feature of AWS Cloud offers the ability to innovate faster and rapidly develop,
test, and launch software applications?
Ability to deploy globally in minutes
Elasticity
Cost savings
Correct answer
Agility
Question 27Skipped
A financial services company is deploying AI systems on AWS to analyze customer
transactions and detect fraud. To meet stringent regulatory requirements, the
company's compliance team needs a tool that can continuously audit AWS usage,
automate evidence collection, and streamline risk assessments. This tool should help
ensure that the AI systems comply with industry standards and reduce the manual
effort involved in compliance reporting.
Which AWS tool meets these requirements?
AWS Trusted Advisor
AWS Artifact
AWS CloudTrail
Correct answer
AWS Audit Manager

Question 28Skipped
Match the following AWS services to the respective use cases:
A) Amazon Textract
B) Amazon Forecast
C) Amazon Kendra
1) Easy-to-use enterprise search service that’s powered by machine learning
2) Automatically extract printed text, handwriting, layout elements, and data from any
document
3) Forecast business outcomes easily and accurately using machine learning
A-1, B-3, C-2
Correct answer
A-2, B-3, C-1
A-3, B-2, C-1
A-2, B-1, C-3

Question 29Skipped
A telecom company is seeking to improve the efficiency and effectiveness of its
customer service operations by integrating generative AI. The goal is to equip customer
service agents with AI-driven tools that can assist in generating accurate, context-aware
responses to customer inquiries, offer real-time suggestions, and help automate routine
tasks. The company is evaluating several generative AI solutions to determine which
one best fits their need for enhancing customer service interactions.
Which of the following is the best fit for this use case?
Amazon Q Developer
Amazon Q Business
Correct answer
Amazon Q in Connect
Amazon Q in QuickSight

Question 30Skipped
A financial services company is building a machine learning model to improve its credit
risk assessment process. The data science team is focused on refining the model’s
inputs to enhance accuracy and performance. To do this, they are exploring the concept
of Feature Engineering, which is crucial for the team to optimize the model’s
predictions.
What is Feature Engineering in the context of machine learning?
Feature Engineering refers to the visualization of data to understand patterns, and it is
important because it helps in identifying trends in the dataset
Feature Engineering is the process of tuning hyperparameters in a machine learning
model, and it is important because it optimizes the model’s performance
Correct answer
Feature Engineering involves selecting, modifying, or creating features from raw data
to improve the performance of machine learning models, and it is important because it
can significantly enhance model accuracy and efficiency
Feature Engineering is the process of collecting raw data, and it is important because
it ensures the availability of data for model training

Question 31Skipped
An app developer is building an educational application to help high-school students
understand fundamental concepts in mathematics, such as calculating the probability
of drawing a spade from a deck of cards.
Which approach would be the most suitable for this purpose?
The developer should utilize unsupervised learning, a machine learning method that
identifies patterns and structures in data without using labeled datasets
The developer should leverage reinforcement learning (RL), a type of machine learning
where an agent learns to make decisions by receiving rewards or penalties
The developer should apply supervised learning, a machine learning approach where
the model learns from labeled datasets to make predictions
Correct answer
The developer should create a rule-based application that uses predefined
mathematical rules and formulas to answer probability questions accurately

Question 32Skipped
Consider a scenario where a fully-managed AWS service needs to be used for
automating the extraction of insights from legal briefs such as contracts and court
records.
What do you recommend?
Amazon Rekognition
Amazon Transcribe
Correct answer
Amazon Comprehend
Amazon Translate

Question 33Skipped
A tech company is building machine learning models using Amazon SageMaker Studio
and wants to streamline its development process. The data science team prefers using
familiar Integrated Development Environments (IDEs) to write, test, and debug code
more efficiently. To ensure a smooth workflow, the team is exploring which IDEs are
supported within SageMaker Studio to maximize productivity and compatibility with
their existing tools.
Which of the following address the given requirement?
RStudio
JupyterLab
Correct answer
All
Code Editor

Question 34Skipped
A company is using a Large Language Model (LLM) on Amazon Bedrock and it wants to
regulate the creativity of the model's output.
Which of the following inference parameters would you recommend for the given use
case?
Correct answer
Temperature
Top P
Top K
Stop sequences

Question 35Skipped
A technology company is considering using Amazon Web Services (AWS) to support its
growing application infrastructure and is exploring different cloud computing models.
The team is particularly interested in Amazon Elastic Compute Cloud (EC2) to handle its
scalable compute needs, as understanding this will help them determine how much
control they have over the underlying infrastructure and how best to manage their
resources.
Which type of cloud computing does Amazon Elastic Compute Cloud (EC2) represent?
Software as a Service (SaaS)
Correct answer
Infrastructure as a Service (IaaS)
Network as a Service (NaaS)
Platform as a Service (PaaS)

Question 36Skipped
A healthcare company is building multiple machine learning models using Amazon
SageMaker to support various projects, such as patient outcome prediction and
medical image analysis. As the number of models grows, the company needs a tool that
provides a centralized view of all models created across its AWS account to easily
track, manage, and monitor them. This will help streamline model governance and
improve operational efficiency.
Which of the following is the best-fit for the given requirements?
Amazon SageMaker Model Monitor
Correct answer
Amazon SageMaker Model Dashboard
Amazon SageMaker Ground Truth
Amazon SageMaker Clarify

Question 37Skipped
A retail company is developing machine learning models on AWS to improve product
recommendations and customer insights. To ensure consistency and collaboration
among its data science team, the company needs a solution for storing, sharing, and
managing the inputs used during the model training and inference phases. The
company is evaluating AWS services that can help streamline this process.
What do you suggest
Amazon SageMaker Ground Truth
Amazon SageMaker Clarify
Amazon SageMaker Data Wrangler
Correct answer
Amazon SageMaker Feature Store

Question 38Skipped
A healthcare company is developing a machine learning model to analyze medical
images and patient records to assist with diagnostics. The team has access to a large
amount of unlabeled data and a smaller set of labeled data, and they are considering
using semi-supervised learning to maximize the utility of both datasets. To make an
informed decision on the approach, the data science team wants to understand which
methods fall under semi-supervised learning.
Which of the following are examples of semi-supervised learning? (Select two)
Correct selection
Sentiment analysis
Dimensionality reduction
Clustering
Correct selection
Fraud identification
Neural network

Question 39Skipped
A research-focused AI company is developing a suite of machine learning models for
tasks such as classification and content generation. The data science team needs to
choose between discriminative and generative models depending on the specific use
case. To make the right decision, they need to understand the fundamental differences
between these two types of models, particularly in the context of generative AI, and how
each model type fits into their project goals.
What is the primary distinction between discriminative models and generative models in
the context of generative AI?
Discriminative models are only used for text classification, while generative models
are only used for image classification
Discriminative models are used to generate new data, while generative models are
used only for classification
Correct answer
Generative models focus on generating new data from learned patterns, whereas
discriminative models classify data by distinguishing between different classes
Generative models are trained on labeled data, while discriminative models can be
trained on both labeled and unlabeled data

Question 40Skipped
A company has implemented a chatbot powered by Amazon Bedrock to handle
customer inquiries and support requests. While the chatbot is effective at providing
automated responses, the company has noticed that some of the replies do not
consistently match its desired tone — professional, empathetic, and friendly. To
maintain brand consistency and ensure a positive customer experience, the company
needs to align the chatbot’s responses with its specific communication style and
standards.
Which approach would be most effective for ensuring that the chatbot's responses are
consistently aligned with the company's tone and style?
The company should adjust the temperature to reduce the randomness and creativity
of the chatbot’s responses
The company should set a low limit on the number of tokens, which restricts the length
of the chatbot's responses, thereby making the chatbot’s responses consistent

Correct answer
The company should iteratively test and adjust the chatbot prompts to ensure that its
outputs consistently reflect the company's tone and style
The company should use batch inferencing, a method that processes multiple input
requests in one go to make the chatbot’s responses consistent

Question 41Skipped
A technology consulting firm is advising a client on the use of AI to enhance their
business operations, particularly through the implementation of large-scale models that
can handle diverse tasks such as text generation, image recognition, and natural
language understanding. The firm is evaluating Foundation Models as a potential
solution and wants to clarify their capabilities, including their ability to generalize across
multiple domains and perform a wide range of tasks with minimal fine-tuning.
Which of the following options aptly summarizes the capabilities of Foundation
Models?
Foundation models can only perform a single task they were specifically trained for
Correct answer
Foundation models can perform a wide range of tasks across different domains by
leveraging their extensive pre-training on large datasets
Foundation models are limited to simple data processing tasks and cannot handle
complex operations
Foundation models are designed to work exclusively with structured data and cannot
process unstructured data like text or images

Question 42Skipped
A logistics company is exploring ways to label large datasets for an upcoming machine
learning project focused on optimizing delivery routes. The team is evaluating two AWS
services—Amazon Mechanical Turk and Amazon Ground Truth—to assist with the data
labeling process. They need to understand the key differences between the two
services, particularly in terms of automation, scalability, and workforce management.
What is the primary difference between Amazon Mechanical Turk and Amazon Ground
Truth?
Correct answer
Amazon Mechanical Turk provides a marketplace for outsourcing various tasks to a
distributed workforce, while Amazon Ground Truth is specifically designed for creating
labeled datasets for machine learning, incorporating both automated and human
labeling
Amazon Mechanical Turk is exclusively for data labeling tasks, whereas Amazon
Ground Truth supports a wide range of tasks including surveys and content
moderation
Amazon Mechanical Turk is used for creating labeled datasets using automated
processes, whereas Amazon Ground Truth is a marketplace for outsourcing various
tasks to a distributed workforce
Amazon Mechanical Turk and Amazon Ground Truth are the same service, used
interchangeably for any task involving human intelligence
Question 43Skipped
A retail company is building a machine learning model to forecast demand for its
products, but the data science team is facing challenges in balancing model complexity
and accuracy. They are trying to avoid overfitting as well as underfitting, since
understanding the differences between these two issues is crucial for optimizing the
model's performance on both historical and unseen data.
How would you differentiate between overfitting and underfitting in the context of
machine learning?
Overfitting is desirable as it ensures the model captures all nuances in the training
data, while underfitting is desirable as it ensures the model generalizes well to new
data
Overfitting and underfitting both refer to a model performing equally well on both the
training data and new, unseen data
Correct answer
Overfitting occurs when a model performs well on the training data but poorly on new,
unseen data, while underfitting occurs when a model performs poorly on both the
training data and new, unseen data
Overfitting occurs when a model is too simple to capture the underlying patterns in the
data, while underfitting occurs when a model is too complex and captures noise rather
than the actual patterns

Question 44Skipped
A company is using Amazon Bedrock and it wants to set an upper limit on the number
of tokens returned in the model's response.
Which of the following inference parameters would you recommend for the given use
case?
Stop sequence
Top K
Correct answer
Response length
Top P

Question 45Skipped
A media company is looking to enhance its content creation processes by leveraging
cutting-edge technologies and has been exploring the use of generative AI. The
leadership team wants to understand the broader significance of generative AI in
modern technological applications, including its potential to automate tasks, improve
creativity, and optimize workflows across industries. Understanding why generative AI
is considered important will help the company make informed decisions about
integrating this technology into its operations.
Given this context, why is generative AI considered important in modern technological
applications?
Generative AI can replace all traditional databases with its own storage solutions
Generative AI can easily perform simple tasks like sorting and filtering data
Generative AI is the best fit for use cases related to gaming and entertainment
applications
Correct answer
Generative AI is important because it can autonomously create novel and complex
data, enhancing creativity and efficiency in various domains

Question 46Skipped
A healthcare analytics company is developing a machine learning model to provide
patient risk assessments based on incoming medical data. The team needs to deploy
the model in a way that supports continuous, low-latency predictions, where each
request receives a response immediately, such as during patient check-ins. The team is
evaluating deployment options within Amazon SageMaker and needs a solution that
offers persistent endpoints to handle these individual prediction requests.
What do you recommend?
Correct answer
Real-time hosting services
Batch transform
Asynchronous Inference
Serverless Inference

Question 47Skipped
A wildlife research organization has gathered thousands of images from camera traps
in natural reserves worldwide, capturing various animal species. To support their
research and conservation efforts, the organization wants to build a system that can
automatically identify and categorize each animal in these images accurately and
efficiently. The organization is evaluating different AI and machine learning techniques
to achieve this goal.
Which approach would you suggest to effectively recognize and categorize the various
animal species in their image dataset?
The company should use face recognition, a computer vision technique specifically
designed to identify faces in images or videos

Correct answer
The company should use object detection, which involves identifying and locating
specific objects within an image
The company should use thermal imaging, a technique that detects heat patterns and
variations, which can identify living beings in low-visibility conditions
The company should use named entity recognition, a technique for identifying entities
like names, places, or dates

Question 48Skipped
A company stores its training datasets on Amazon S3 in the form of tabular data
running into millions of rows. The company needs to prepare this data for Machine
Learning jobs. The data preparation involves data selection, cleansing, exploration, and
visualization using a single visual interface.
Which Amazon SageMaker service is the best fit for these requirements?
SageMaker Model Dashboard
Amazon SageMaker Feature Store
Correct answer
Amazon SageMaker Data Wrangler
Amazon SageMaker Clarify

Question 49Skipped
The development team at a company needs to select the most appropriate large
language model (LLM) for the company's flagship application. Given the vast array of
LLMs available, the team is uncertain about the best choice. Additionally, since the
application will be publicly accessible, the team has concerns about the possibility of
generating harmful or inappropriate content.
Which AWS solutions should the team implement to address both the selection of the
appropriate model and the mitigation of harmful content generation? (Select two).
Amazon Comprehend
Correct selection
Model Evaluation on Amazon Bedrock
Correct selection
Guardrails for Amazon Bedrock
Amazon SageMaker Model Monitor
Amazon SageMaker Clarify

Question 50Skipped
A growing e-commerce company is considering migrating its infrastructure to the cloud
to improve operational efficiency and scalability. The leadership team is evaluating the
benefits of cloud computing, such as cost savings, flexibility, and enhanced
collaboration, but wants to understand the specific advantages that cloud services can
offer over traditional on-premises infrastructure.
Which of the following are the advantages of cloud computing? (Select three)
Spend money on building and maintaining data centers
Trade variable expense for capital expense
Correct selection
Benefit from massive economies of scale
Correct selection
Go global in minutes and deploy applications in multiple regions around the world with
just a few clicks
Allocate a few months of planning for your infrastructure capacity needs
Correct selection
Trade capital expense for variable expense

Question 51Skipped
A financial services company is exploring Amazon Bedrock to streamline its AI
development for use cases such as fraud detection, personalized customer service, and
automated reporting. The company is particularly interested in understanding the key
features and benefits of Amazon Bedrock, including its ability to simplify access to
powerful foundation models, support customizations, and integrate with existing AWS
services.
To make an informed decision, the company needs to identify which of the following
accurately applies to Amazon Bedrock and its capabilities? (Select two)
Correct selection
You can use a customized model only in the Provisioned Throughput mode
You can use a customized model in the Provisioned Throughput or On-Demand mode
Larger models are cheaper to use than smaller models
You can use the On-Demand mode only with time-based term commitments
Correct selection
Smaller models are cheaper to use than larger models

Question 52Skipped
A company developing AI-powered customer service chatbots is exploring ways to
improve the quality and accuracy of responses using Reinforcement Learning from
Human Feedback (RLHF). The data science team is considering using Amazon
SageMaker Ground Truth to assist with gathering and processing human feedback
during model training. To ensure this solution aligns with their needs, they want to
understand how SageMaker Ground Truth supports the key capabilities required for
implementing RLHF, such as collecting, labeling, and managing human input effectively.
What do you suggest?
SageMaker Ground Truth automatically generates synthetic data for training
reinforcement learning models without any human intervention
SageMaker Ground Truth uses pre-trained models to eliminate the need for human
feedback in the reinforcement learning process
Correct answer
SageMaker Ground Truth enables the creation of high-quality labeled datasets by
incorporating human feedback in the labeling process, which can be used to improve
reinforcement learning models
SageMaker Ground Truth is specifically designed for real-time decision-making in
autonomous systems, bypassing the need for any data labeling

Question 53Skipped
A large enterprise is looking to implement an AI-powered assistant to help employees
across departments streamline their work by answering questions, summarizing
reports, generating content, and securely accessing data from internal systems. The
company needs a solution that can seamlessly integrate with its enterprise systems
while ensuring data privacy and security. The team is exploring various generative AI-
powered assistants that can fulfill these requirements.
Which of the following is a generative AI–powered assistant that can answer questions,
provide summaries, generate content, and securely complete tasks based on data and
information in the enterprise systems?
Amazon Q in Connect
Correct answer
Amazon Q Business
Amazon Q in QuickSight
Amazon Q Developer

Question 54Skipped
A healthcare company is deploying AI models using Amazon SageMaker to predict
patient outcomes and ensure compliance with healthcare regulations. The data science
team wants to document important details about their models, such as performance,
bias assessments, and intended use. They are considering using SageMaker model
cards for this purpose but also want to understand how AI service cards fit into the
broader documentation of their AI services. Understanding the differences between
these two tools will help the team select the right one for tracking and managing their AI
models.
Given this context, how would you highlight the key differences between SageMaker
model cards and AI service cards?
SageMaker model cards are used to store data for machine learning models, while AI
service cards are used for storing user credentials
SageMaker model cards are used exclusively for monitoring model performance,
whereas AI service cards are used for managing model security
Correct answer
SageMaker model cards include information about the model such as intended use
and risk rating of a model, training details and metrics, evaluation results, and
observations. AI service cards provide transparency about AWS AI services' intended
use, limitations, and potential impacts
SageMaker model cards provide technical documentation for deploying models, while
AI service cards offer transparency about the intended use, limitations, and potential
impacts of AWS AI services

Question 55Skipped
A financial services company is deploying machine learning models to automate fraud
detection but wants to ensure continuous model accuracy and compliance with
regulatory standards. The data science team is exploring AWS services that can help in
monitoring machine learning models and incorporating human review processes.
Understanding which AWS services are specifically designed to support model
monitoring and human oversight will help the team maintain high standards of accuracy
and compliance.
Which AWS services can be combined to support these requirements? (Select two)
Amazon SageMaker Data Wrangler
Amazon SageMaker Feature Store
Amazon SageMaker Ground Truth
Correct selection
Amazon Augmented AI (Amazon A2I)
Correct selection
Amazon SageMaker Model Monitor

Question 56Skipped
A machine learning team at a tech company is developing a generative AI model to
automate text generation for customer support. As part of optimizing the model’s
performance, the team needs to adjust both model parameters and hyperparameters
but wants to clearly understand the distinctions between the two. Understanding these
differences is crucial for fine-tuning the model and improving its output.
Which of the following highlights the key differences between model parameters and
hyperparameters in the context of generative AI?
Both Hyperparameters and model parameters are values that can be adjusted for
model customization to control the training process

Correct answer
Model parameters are values that define a model and its behavior in interpreting input
and generating responses. Hyperparameters are values that can be adjusted for model
customization to control the training process
Hyperparameters are values that define a model and its behavior in interpreting input
and generating responses. Model parameters are values that can be adjusted for
model customization to control the training process
Both Hyperparameters and model parameters are values that define a model and its
behavior in interpreting input and generating responses
Question 57Skipped
Consider the following AI model responses to user prompts. Which response
exemplifies hallucination and which exemplifies toxicity?
Prompt 1: "What is the capital of France?"
Response A: "The capital of France is Mars."
Prompt 2: "Tell me a joke."
Response B: "Why don't scientists trust atoms? Because they make up everything!"
Prompt 3: "Describe a good book for children."
Response C: "You should read 'ABC', a great book with fun activities for kids."
Prompt 4: "What do you think about people from [specific group]?"
Response D: "People from [specific group] are inferior and should not be trusted."
Response C is hallucination; Response B is toxicity
Response B is hallucination; Response C is toxicity
Correct answer
Response A is hallucination; Response D is toxicity
Response D is hallucination; Response A is toxicity

Question 58Skipped
An Internet-of-Things (IoT) company is developing a suite of smart sensors and devices
that rely on real-time data processing to enable applications like predictive
maintenance, environmental monitoring, and immediate anomaly detection. To provide
immediate feedback and actions, the company needs to deploy machine learning
models directly on its edge devices, ensuring that these models can perform inference
with minimal latency. The company is evaluating different approaches to optimize
performance and maintain low-latency inference on these edge devices.
Which approach would be the most suitable for meeting this requirement?
The company should use a central API connected to a large language model (LLM)
with an asynchronous inference endpoint, which allows the model to handle requests
from multiple edge devices
The company should use a central API connected to a small language model (SLM)
with an asynchronous inference endpoint, which allows the model to handle requests
from multiple edge devices
Correct answer
The company should use an optimized small language model (SLM) deployed directly
on the edge device, allowing for real-time, low-latency inference
The company should use an optimized large language model (LLM) deployed directly
on the edge device, allowing for real-time, low-latency inference

Question 59Skipped
A retail company is developing a machine learning model to predict customer churn and
is in the process of preparing its dataset. The data science team plans to divide the
data into a training set, validation set, and test set to ensure the model performs well
across different stages of development and evaluation. To proceed effectively, the team
needs to fully understand the roles of each of these sets and how they contribute to
building a robust model.
Which of the following is correct regarding the training set, validation set, and test set
used in the context of machine learning? (Select two)
Test sets are optional
Validation set is used to determine how well the model generalizes
Test set is used for hyperparameter tuning
Correct selection
Test set is used to determine how well the model generalizes
Correct selection
Validation sets are optional

Question 60Skipped
A financial services company is building a machine learning model to predict loan
defaults, but the data science team is struggling to find the right balance between
model complexity and accuracy. They are aware of the bias-variance trade-off, as
understanding this trade-off is critical for optimizing the model’s performance and
ensuring it generalizes well.
What is the bias versus variance trade-off in machine learning?
Correct answer
The bias versus variance trade-off refers to the challenge of balancing the error due to
the model's complexity (variance) and the error due to incorrect assumptions in the
model (bias), where high bias can cause underfitting and high variance can cause
overfitting
The bias versus variance trade-off refers to the balance between underfitting and
overfitting, where high bias leads to overfitting and high variance leads to underfitting
The bias versus variance trade-off involves choosing between a model with high
complexity that may capture more noise (high bias) and a simpler model that may
generalize better but miss important patterns (high variance)
The bias versus variance trade-off is a technique used to improve model performance
by increasing both bias and variance simultaneously to achieve better generalization

Question 61Skipped
A healthcare analytics company is exploring the use of Foundation Models to automate
the process of labeling vast amounts of medical data, such as patient records and
clinical notes, to enhance its machine learning models for diagnosis and treatment
recommendations. The company wants to understand the specific techniques that
Foundation Models use to generate labels from raw input data, helping streamline the
data annotation process without requiring extensive manual effort.
Which of the following techniques is used by Foundation Models to create labels from
input data?
Supervised learning
Correct answer
Self-supervised learning
Reinforcement learning
Unsupervised learning

Question 62Skipped
A financial services company is developing machine learning models using Amazon
SageMaker to automate loan approvals and fraud detection. To ensure transparency
and compliance with industry regulations, the data science team needs a tool that
documents the model’s intended uses, performance metrics, and any assumptions
made during development. This documentation is crucial for maintaining accountability
and ensuring that the models are used appropriately.
Which Amazon SageMaker tool meets these requirements?
Amazon SageMaker Clarify
Amazon SageMaker Canvas
Correct answer
Amazon SageMaker Model Cards
Amazon SageMaker Model Monitor

Question 63Skipped
A financial services company is deploying AI models to assess credit risk and make
lending decisions. As part of ensuring ethical AI use, the company wants to build
models that are both interpretable and explainable to regulators, stakeholders, and
customers. The data science team needs to understand the distinction between
interpretability and explainability in the context of Responsible AI to choose the right
techniques for transparency. This distinction will guide the company in making its AI
models more trustworthy and compliant.
Which of the following represents the best option for the given use case?
Interpretability is used to enhance the model's performance, while explainability is
used to ensure the model's security
Interpretability refers to the ability to understand the technical details of the model's
code, while explainability refers to the ability to reproduce the model's results
Explainability is about understanding the internal mechanisms of a machine learning
model, whereas interpretability focuses on providing understandable reasons for the
model's predictions and behaviors to stakeholders
Correct answer
Interpretability is about understanding the internal mechanisms of a machine learning
model, whereas explainability focuses on providing understandable reasons for the
model's predictions and behaviors to stakeholders

Question 64Skipped
A software development team is exploring tools to improve their coding efficiency and
streamline their workflow. They are particularly interested in leveraging Amazon Q
Developer to support their development processes, but they want to understand its
specific capabilities.
Which of the following accurately describes what Amazon Q Developer can do to
enhance the team's development efforts?
Correct answer
Amazon Q Developer can suggest code snippets, providing developers with
recommendations for code based on specific tasks or requirements
Amazon Q Developer can deploy applications, automating the entire process of
application deployment from development to production environments
Amazon Q Developer can create Large Language Model (LLM) chatbots, enabling the
design and deployment of large language model-based conversational agents
Amazon Q Developer can create SageMaker models, allowing users to develop, train,
and deploy machine learning models within the Amazon SageMaker environment
Question 65Skipped
A financial services firm is adopting Amazon Q Business to streamline its data-driven
decision-making processes. As part of the implementation, the company needs a robust
solution for managing user access, ensuring that employees across various
departments have appropriate permissions to interact with dashboards and reports.
The team is evaluating options for user management that offer secure, scalable, and
easy-to-administer controls within Amazon Q Business.
Which of the following would you recommend for user management in Amazon Q
Business?
IAM user
AWS Account
AWS IAM service
Correct answer
IAM Identity Center

Question 66Skipped
A financial services company is developing machine learning models to improve fraud
detection and credit risk assessment. However, given the complexity of these tasks, the
team wants to incorporate human input at key stages of the machine learning lifecycle
to ensure that the models are accurate and relevant. The company is looking for an
AWS service that allows human input and feedback to be integrated into the model
development process, improving the overall performance and trustworthiness of the
models.
What do you suggest?
Correct answer
Amazon SageMaker Ground Truth
Amazon SageMaker Role Manager
Amazon SageMaker Clarify
Amazon SageMaker Feature Store

Question 67Skipped
An e-learning company is developing a Large Language Model (LLM) chatbot using
Amazon Bedrock to enhance the personalized learning experience on its platform. The
chatbot needs to dynamically tailor its responses based on the user's age group. By
leveraging Amazon Bedrock's foundation models, the company aims to create an
adaptive learning tool that delivers relevant, engaging, and age-appropriate support to a
diverse user base.
As an AI Practitioner, which of the following solutions would you recommend?
Perform model re-training for tailoring responses based on user age
Perform fine-tuning for the model to adjust the style or tone of responses based on
user age
Correct answer
Implement dynamic prompt engineering to customize responses based on user
characteristics like age
Leverage Retrieval-Augmented Generation (RAG) to customize responses based on
user characteristics like age

Question 68Skipped
A media company is developing generative AI applications on AWS to automate content
creation and enhance customer engagement. Given the sensitivity of customer data
and the complexity of AI models, the company’s security team wants to implement a
defense-in-depth security approach to protect both the data and the AI infrastructure.
Which of the following strategies best aligns with the given requirements?
Relying solely on data encryption to protect the AI training data
Correct answer
Applying multiple layers of security measures including input validation, access
controls, and continuous monitoring to address vulnerabilities
Using a single authentication mechanism for all users and services accessing the AI
models
Implementing a single-layer firewall to block unauthorized access to the AI models

Question 69Skipped
A software company is developing a generative AI model for language translation and
needs to optimize the way the model processes and understands text. The development
team is focusing on improving the model’s ability to convert words into a form that the
AI can effectively interpret and generate accurate translations. To achieve this, they
need to clarify the roles of tokens and embeddings in the model’s language processing.
Which of the following summarizes the differences between a token and an embedding
in the context of generative AI?
Both token and embedding refer to a sequence of characters that a model can
interpret or predict as a single unit of meaning
Correct answer
A token is a sequence of characters that a model can interpret or predict as a single
unit of meaning, whereas, an embedding is a vector of numerical values that
represents condensed information obtained by transforming input into that vector
Both token and embedding refer to a vector of numerical values that represents
condensed information obtained by transforming input into that vector
An embedding is a sequence of characters that a model can interpret or predict as a
single unit of meaning, whereas, a token is a vector of numerical values that
represents condensed information obtained by transforming input into that vector

Question 70Skipped
A financial services company is developing a Deep Learning model to detect fraudulent
transactions in real-time. The data science team has decided to use neural networks as
the backbone of the model but needs to fully understand how neural networks function,
as understanding the working principles of neural networks is crucial for building an
effective fraud detection system.
How do neural networks work in the context of Deep Learning?
Neural networks operate by storing all possible outcomes and selecting the most
appropriate one for each input
Neural networks rely solely on predefined mathematical formulas and do not learn
from data
Neural networks learn to perform tasks by being explicitly programmed with rules for
each task
Correct answer
Neural networks consist of layers of nodes (neurons) that process input data,
adjusting the weights of connections between nodes through training to recognize
patterns and make predictions

Question 71Skipped
A social media company is implementing an AI-driven content recommendation system
to enhance user engagement. During testing, the data science team notices that the AI
suggests content differently based on user demographics, leading to concerns about
whether the model is treating all users fairly. The team wants to ensure the system
avoids any form of bias and complies with ethical AI standards. To better understand
this issue, they need a clear example of algorithmic bias to recognize and address it in
their system.
Which of the following scenarios best illustrates algorithmic bias?
A customer service representative resolves complaints based on their judgment rather
than company policy
A human resources manager hires candidates based on personal interviews without
considering their resumes
Correct answer
A hiring algorithm consistently prefers candidates from a particular gender, even
though the candidates' qualifications are similar across genders
A weather prediction model occasionally makes incorrect forecasts due to random
fluctuations in weather patterns

Question 72Skipped
A legal firm is looking to implement an AI solution that can generate detailed, accurate
responses to client queries by retrieving relevant information from its extensive
database of legal documents. The firm is considering the use of Retrieval Augmented
Generation (RAG) through Amazon Bedrock to enhance the quality and relevance of the
generated content. The team wants to understand the best-fit use cases for RAG to
determine if it aligns with their needs for knowledge retrieval and content generation.
Which of the following represents the best-fit use cases for utilizing Retrieval
Augmented Generation (RAG) in Amazon Bedrock? (Select two)
Correct selection
Customer service chatbot
Original content creation
Image generation from text prompt
Correct selection
Medical queries chatbot
Product recommendations that match shopper preferences

Question 73Skipped
A business needs an automated solution that can extract text from thousands of
receipts and invoices generated across all its stores.
Which AWS Machine Learning (ML) service can offer the most optimal solution for this
use case ?
Correct answer
Amazon Textract
Amazon Comprehend
Amazon Transcribe
Amazon Rekognition

Question 74Skipped
A retail company is embarking on a machine learning project to enhance customer
segmentation and personalize marketing campaigns. As the data science team begins
planning the implementation, the team wants to identify the primary challenges in
machine learning implementation. Understanding these challenges will help the team
anticipate potential roadblocks and develop strategies to overcome them.
Which of the following represents the best option for the given use case?
Correct answer
Difficulty in collecting and preparing high-quality data for training models
Limited applications of machine learning in real-world scenarios
Lack of available machine learning algorithms
Insufficient computational power to run basic machine learning models

Question 75Skipped
A global e-commerce company is leveraging a Foundation Model (FM) to improve its
product recommendation engine and enhance customer experience. However, the data
science team is looking to further optimize the model's performance by applying
advanced techniques that can fine-tune the FM for specific tasks, ensure higher
accuracy, and improve overall efficiency. The company needs to identify the most
effective methods for enhancing the model's capabilities while maintaining scalability.
Which of the following is correct regarding the techniques used to improve the
performance of a Foundation Model (FM)?
Neither Fine-tuning nor Retrieval-augmented generation (RAG) changes the weights of
the FM
Fine-tuning does not change the weights of the FM whereas Retrieval-augmented
generation (RAG) changes the weights of the FM
Both Fine-tuning and Retrieval-augmented generation (RAG) change the weights of
the FM
Correct answer
Fine-tuning changes the weights of the FM whereas Retrieval-augmented generation
(RAG) does not change the weights of the FM

Question 76Skipped
A healthcare company is considering migrating its on-premises infrastructure to AWS
Cloud to enhance data management, improve scalability, and reduce operational costs.
The leadership team is new to cloud technologies and wants a clear understanding of
what cloud computing entails, specifically how AWS defines and implements it. This
understanding is essential for the company to make informed decisions about the
benefits and use cases of cloud computing in their operations.
Given this context, what is cloud computing, as defined by AWS?
Cloud computing is the process of using a single local server to store and process
data
Correct answer
Cloud computing refers to the on-demand delivery of IT resources and applications via
the internet with pay-as-you-go pricing
Cloud computing is the practice of using only open-source software for all computing
needs
Cloud computing involves manually managing physical data centers and networking
hardware for data storage and processing

Question 77Skipped
A financial services company is developing a machine-learning model to classify loan
applications as either "approved" or "denied." To ensure the model performs effectively,
the company wants to evaluate how accurately it predicts these outcomes. Specifically,
they are interested in knowing the overall percentage of correct predictions, including
both approved and denied applications. The company is considering several metrics to
assess the model's performance in terms of the number of correct outcomes.
Which metric would be most appropriate for this purpose?

Correct answer
The company should use Accuracy, which measures the proportion of correctly
predicted instances (both true positives and true negatives) out of the total number of
instances
The company should use F1 Score, a metric that considers both precision and recall by
calculating their harmonic mean
The company should use Root Mean Squared Error (RMSE), a metric that calculates
the square root of the average of the squared differences between predicted and
actual values
The company should use R-squared, a statistical measure that indicates the
proportion of variance in the dependent variable explained by the independent
variables

Question 78Skipped
A technology firm is developing an AI-driven solution for automating business
processes and needs to design effective prompts for its generative AI model. The
model is tasked with solving complex, multi-step problems, such as generating detailed
business reports or creating process workflows. To improve the model's performance,
the team is exploring prompt engineering techniques that can help simplify these tasks
by breaking them down into smaller, manageable parts.
Which prompt engineering technique is best suited for breaking down a complex
problem into smaller logical parts?
Few shot Prompting
Correct answer
Chain-of-thought prompting
Zero shot Prompting
Negative prompting

Question 79Skipped
A software development company is exploring Amazon Q Developer to enhance its
internal tools and workflows. The company is particularly interested in leveraging the
platform's capabilities to automate code generation, improve task automation, and
integrate machine learning features into its applications. To understand how Amazon Q
Developer can support these objectives, the development team needs a clear overview
of its core functionalities.
Which of the following represents the capabilities of Amazon Q Developer? (Select two)
Correct selection
Understand and manage your cloud infrastructure on AWS
Deploy your cloud infrastructure on AWS

Correct selection
Get answers to your AWS account-specific cost-related questions using natural
language
Visualize your AWS account-specific cost-related data in Amazon Q Developer
Modify your AWS resources to achieve cost-optimization

Question 80Skipped
A media company is exploring cutting-edge AI models to automate tasks such as
content generation and language translation. The development team is particularly
interested in using Transformer models due to their efficiency and performance in
natural language processing tasks. To make an informed decision, the team needs to
identify which models belong to the Transformer architecture and how they can be
applied to their use cases.
Which of the following is an example of a Transformer model?
Adobe Firefly
Correct answer
ChatGPT
Stable Diffusion
DALL-E

Question 81Skipped
A retail company is exploring advanced AI solutions to enhance customer experience by
integrating both visual and textual data for tasks such as product recommendations,
automated image tagging, and customer support. The team is considering using
multimodal models, which can process and understand multiple types of input data, but
they need a clear understanding of how these models work and their key advantages.
To help make an informed decision, the company wants to clarify the capabilities of
multimodal models.
Which of the following summarizes the capabilities of a multimodal model?
A multimodal model can accept only a single type of input and it can only create a
single type of output
Correct answer
A multimodal model can accept a mix of input types such as audio/text and create a
mix of output types such as video/image
A multimodal model can accept only a single type of input, however, it can create a
mix of output types such as video/image
A multimodal model can accept a mix of input types such as audio/text, however, it
can only create a single type of output

Question 82Skipped
A company wants to use Amazon Bedrock for several use cases involving text
generation as well as image generation. Which of the following Foundation Models do
you recommend?
Jurassic
Llama
Correct answer
Amazon Titan
Claude

Question 83Skipped
A company is using Amazon Personalize to build a recommendations engine for its e-
commerce application. As part of the process, the data from ten different sources
needs to be processed and imported into Amazon Personalize.
Which AWS service will help import, prepare, and transform data before it is fed into
Amazon Personalize?
Amazon SageMaker Clarify
Amazon SageMaker Ground Truth
Correct answer
Amazon SageMaker Data Wrangler
Amazon SageMaker Feature Store

Question 84Skipped
A robotics company is developing an AI system to improve the autonomous navigation
of its robots. The team is exploring Deep Learning to enhance the system’s ability to
recognize and respond to its environment. To ensure the AI model performs well, the
team needs to understand how model training works in Deep Learning, specifically the
process through which the model learns from large datasets by adjusting its internal
parameters. This understanding is essential to optimize the model for real-time
decision-making.
How does model training work in Deep Learning?
Model training in deep learning requires no data; the neural network automatically
learns from predefined algorithms without any input
Correct answer
Model training in deep learning involves using large datasets to adjust the weights and
biases of a neural network through multiple iterations, using techniques such as
gradient descent to minimize the error
Model training in deep learning involves only the use of support vector machines and
decision trees to create predictive models
Model training in deep learning involves manually setting the weights and biases of a
neural network based on predefined rules

Question 85Skipped
An e-commerce company is developing a chatbot to enhance its user experience by
allowing customers to submit queries that include both text descriptions and images,
such as product photos or screenshots of issues. The company aims for the chatbot to
understand these multi-modal inputs and provide accurate and context-aware
responses, seamlessly combining visual and textual information to address customer
needs effectively.
Which approach would be the most cost-effective for enabling the chatbot to process
such multi-modal queries effectively?
The company should use a text-only language model, which is trained exclusively on
textual data
The company should use a multi-modal generative model, which can generate
responses or outputs based on combined inputs from different modalities, such as
text and images, enhancing the chatbot’s ability to provide contextually relevant
answers
Correct answer
The company should use a multi-modal embedding model, which is designed to
represent and align different types of data (such as text and images) in a shared
embedding space, allowing the chatbot to understand and interpret both forms of input
simultaneously
The company should use a convolutional neural network (CNN), a deep learning model
primarily designed for processing image data