2026 AWS Certified AI Practitioner (AIF-C01) Exam
with All 65 Questions, with 100% Correct Answers
and Rationale
DOMAIN 1: FUNDAMENTALS OF AI AND ML (Questions 1-13)
Question 1
A company is developing a machine learning model using Amazon SageMaker and needs a
solution to store and share feature sets across different teams for collaborative model building.
Which Amazon SageMaker feature should the company use?
A. Amazon SageMaker Feature Store
B. Amazon SageMaker Data Wrangler
C. Amazon SageMaker Clarify
D. Amazon SageMaker Model Registry
Correct Answer: A
Rationale: SageMaker Feature Store is designed to allow teams to store, manage, and share
features (attributes or variables) in a central repository. This ensures consistency across models
and helps teams collaborate by reusing the same features across multiple projects .
• Amazon SageMaker Data Wrangler is used for data transformation and preparation,
helping users clean and structure data before training models.
• Amazon SageMaker Clarify focuses on detecting bias in machine learning models and
ensuring explainability.
• Amazon SageMaker Model Registry is designed for managing and versioning machine
learning models, not for storing or sharing feature sets.
Question 2
A business uses Amazon SageMaker to run its machine learning pipeline in a production
environment. The company processes large datasets, sometimes reaching 1 GB in size, with
processing times that can take up to an hour. To support its operations, the company requires
low-latency predictions. Which Amazon SageMaker inference option should the company
choose?
A. Real-time inference
B. Serverless inference
C. Asynchronous inference
D. Batch transform
Correct Answer: A
Rationale: Real-time inference is designed for scenarios where low-latency responses are
needed. It is ideal when predictions need to be generated immediately upon receiving input data,
making it suitable for use cases requiring near real-time results, even with large datasets .
• Serverless inference is optimized for intermittent workloads that don't require low-
latency predictions.
• Asynchronous inference is intended for situations where the input size is large or the
processing time is lengthy, but real-time predictions are not required.
• Batch transform is used for processing large datasets in batches without a focus on real-
, time results.
Question 3
A data scientist needs to convert spoken language from customer service calls into text for
analysis. Which AWS service should they use?
A. Amazon Polly
B. Amazon Transcribe
C. Amazon Translate
D. Amazon Comprehend
Correct Answer: B
Rationale: Amazon Transcribe is an AWS service that provides speech-to-text capabilities,
converting audio files into text. This is ideal for analyzing customer service calls .
• Amazon Polly converts text to speech, not speech to text.
• Amazon Translate provides language translation services.
• Amazon Comprehend performs natural language processing (NLP) to extract insights
from text.
Question 4
Which type of machine learning uses labeled data to train models to predict outcomes or classify
data?
A. Supervised learning
B. Unsupervised learning
C. Reinforcement learning
D. Transfer learning
Correct Answer: A
Rationale: Supervised learning uses labeled datasets with input-output pairs to train models for
tasks like classification and regression .
• Unsupervised learning finds patterns in unlabeled data without predefined labels.
• Reinforcement learning learns through rewards and penalties based on actions.
• Transfer learning adapts pre-trained models to new tasks.
Question 5
A company wants to segment its customers into distinct groups based on purchasing behavior
without any predefined categories. Which type of machine learning should they use?
A. Supervised learning
B. Unsupervised learning
C. Reinforcement learning
D. Active learning
Correct Answer: B
Rationale: Unsupervised learning is used for clustering and finding patterns in data without
predefined labels, making it ideal for customer segmentation tasks .
• Supervised learning requires labeled data, which would not be available for discovering
unknown customer segments.
• Reinforcement learning is used for sequential decision-making, not segmentation.
• Active learning involves the model selectively querying for labels, which still requires
labeled data.
, Question 6
A developer needs to add conversational AI capabilities to their application to create a chatbot.
Which AWS service should they use?
A. Amazon Comprehend
B. Amazon Lex
C. Amazon Polly
D. Amazon Rekognition
Correct Answer: B
Rationale: Amazon Lex is a service for building conversational interfaces using voice and text,
powering chatbots and conversational AI applications .
• Amazon Comprehend performs text analysis and NLP, but not conversational AI.
• Amazon Polly converts text to speech.
• Amazon Rekognition provides image and video analysis.
Question 7
What is the primary purpose of feature engineering in the machine learning lifecycle?
A. To transform raw data into features that better represent the underlying problem to improve
model accuracy
B. To label training data for supervised learning
C. To evaluate model performance on test data
D. To deploy the model to production
Correct Answer: A
Rationale: Feature engineering involves transforming raw data into features that better represent
the underlying problem, which can significantly improve model accuracy and performance .
• Labeling data is a separate data preparation step, not feature engineering.
• Model evaluation occurs after training, using metrics to assess performance.
• Model deployment is the final stage of making the model available for inference.
Question 8
A machine learning model achieves 99% accuracy on training data but only 75% accuracy on
test data. What is this phenomenon called?
A. Underfitting
B. Overfitting
C. Bias
D. Variance
Correct Answer: B
Rationale: Overfitting occurs when a model learns the training data too well, including noise
and random fluctuations, resulting in poor performance on unseen test data .
• Underfitting occurs when a model fails to capture underlying patterns in the data.
• Bias refers to systematic errors in predictions that can lead to unfair outcomes.
• Variance refers to how much the model's predictions change with different training data.
Question 9
Which AWS service provides a fully managed platform for building, training, and deploying
machine learning models?
with All 65 Questions, with 100% Correct Answers
and Rationale
DOMAIN 1: FUNDAMENTALS OF AI AND ML (Questions 1-13)
Question 1
A company is developing a machine learning model using Amazon SageMaker and needs a
solution to store and share feature sets across different teams for collaborative model building.
Which Amazon SageMaker feature should the company use?
A. Amazon SageMaker Feature Store
B. Amazon SageMaker Data Wrangler
C. Amazon SageMaker Clarify
D. Amazon SageMaker Model Registry
Correct Answer: A
Rationale: SageMaker Feature Store is designed to allow teams to store, manage, and share
features (attributes or variables) in a central repository. This ensures consistency across models
and helps teams collaborate by reusing the same features across multiple projects .
• Amazon SageMaker Data Wrangler is used for data transformation and preparation,
helping users clean and structure data before training models.
• Amazon SageMaker Clarify focuses on detecting bias in machine learning models and
ensuring explainability.
• Amazon SageMaker Model Registry is designed for managing and versioning machine
learning models, not for storing or sharing feature sets.
Question 2
A business uses Amazon SageMaker to run its machine learning pipeline in a production
environment. The company processes large datasets, sometimes reaching 1 GB in size, with
processing times that can take up to an hour. To support its operations, the company requires
low-latency predictions. Which Amazon SageMaker inference option should the company
choose?
A. Real-time inference
B. Serverless inference
C. Asynchronous inference
D. Batch transform
Correct Answer: A
Rationale: Real-time inference is designed for scenarios where low-latency responses are
needed. It is ideal when predictions need to be generated immediately upon receiving input data,
making it suitable for use cases requiring near real-time results, even with large datasets .
• Serverless inference is optimized for intermittent workloads that don't require low-
latency predictions.
• Asynchronous inference is intended for situations where the input size is large or the
processing time is lengthy, but real-time predictions are not required.
• Batch transform is used for processing large datasets in batches without a focus on real-
, time results.
Question 3
A data scientist needs to convert spoken language from customer service calls into text for
analysis. Which AWS service should they use?
A. Amazon Polly
B. Amazon Transcribe
C. Amazon Translate
D. Amazon Comprehend
Correct Answer: B
Rationale: Amazon Transcribe is an AWS service that provides speech-to-text capabilities,
converting audio files into text. This is ideal for analyzing customer service calls .
• Amazon Polly converts text to speech, not speech to text.
• Amazon Translate provides language translation services.
• Amazon Comprehend performs natural language processing (NLP) to extract insights
from text.
Question 4
Which type of machine learning uses labeled data to train models to predict outcomes or classify
data?
A. Supervised learning
B. Unsupervised learning
C. Reinforcement learning
D. Transfer learning
Correct Answer: A
Rationale: Supervised learning uses labeled datasets with input-output pairs to train models for
tasks like classification and regression .
• Unsupervised learning finds patterns in unlabeled data without predefined labels.
• Reinforcement learning learns through rewards and penalties based on actions.
• Transfer learning adapts pre-trained models to new tasks.
Question 5
A company wants to segment its customers into distinct groups based on purchasing behavior
without any predefined categories. Which type of machine learning should they use?
A. Supervised learning
B. Unsupervised learning
C. Reinforcement learning
D. Active learning
Correct Answer: B
Rationale: Unsupervised learning is used for clustering and finding patterns in data without
predefined labels, making it ideal for customer segmentation tasks .
• Supervised learning requires labeled data, which would not be available for discovering
unknown customer segments.
• Reinforcement learning is used for sequential decision-making, not segmentation.
• Active learning involves the model selectively querying for labels, which still requires
labeled data.
, Question 6
A developer needs to add conversational AI capabilities to their application to create a chatbot.
Which AWS service should they use?
A. Amazon Comprehend
B. Amazon Lex
C. Amazon Polly
D. Amazon Rekognition
Correct Answer: B
Rationale: Amazon Lex is a service for building conversational interfaces using voice and text,
powering chatbots and conversational AI applications .
• Amazon Comprehend performs text analysis and NLP, but not conversational AI.
• Amazon Polly converts text to speech.
• Amazon Rekognition provides image and video analysis.
Question 7
What is the primary purpose of feature engineering in the machine learning lifecycle?
A. To transform raw data into features that better represent the underlying problem to improve
model accuracy
B. To label training data for supervised learning
C. To evaluate model performance on test data
D. To deploy the model to production
Correct Answer: A
Rationale: Feature engineering involves transforming raw data into features that better represent
the underlying problem, which can significantly improve model accuracy and performance .
• Labeling data is a separate data preparation step, not feature engineering.
• Model evaluation occurs after training, using metrics to assess performance.
• Model deployment is the final stage of making the model available for inference.
Question 8
A machine learning model achieves 99% accuracy on training data but only 75% accuracy on
test data. What is this phenomenon called?
A. Underfitting
B. Overfitting
C. Bias
D. Variance
Correct Answer: B
Rationale: Overfitting occurs when a model learns the training data too well, including noise
and random fluctuations, resulting in poor performance on unseen test data .
• Underfitting occurs when a model fails to capture underlying patterns in the data.
• Bias refers to systematic errors in predictions that can lead to unfair outcomes.
• Variance refers to how much the model's predictions change with different training data.
Question 9
Which AWS service provides a fully managed platform for building, training, and deploying
machine learning models?
