Cert-Pass
Log in Sign up
AWS

AWS Certified AI Practitioner AIF-C01 Practice Questions

1100 exam-accurate questions with explanations

info

Free Sample Questions

Showing 10 of 1100 questions. Get full access to all questions, detailed explanations, and study materials.

1
Domain 1: Fundamentals of AI and ML

[Easy] Matching-style question: A cloud practitioner at a regional bank is validating a study map before an architecture workshop. Which option correctly matches every item to its primary purpose?

A Amazon Transcribe = text to speech; Amazon Polly = speech to text; Amazon Comprehend = sentiment and entity analysis
B Amazon Transcribe = speech to text; Amazon Polly = sentiment and entity analysis; Amazon Comprehend = text to speech
C Amazon Transcribe = sentiment and entity analysis; Amazon Polly = text to speech; Amazon Comprehend = speech to text
D Amazon Transcribe = speech to text; Amazon Polly = text to speech; Amazon Comprehend = sentiment and entity analysis check_circle
lightbulb

Explanation

Option D correctly maps all three items. The distractors swap closely related concepts or services, which is a common exam trap.

2
Domain 1: Fundamentals of AI and ML

[Easy] A governance checkpoint at a payment processor confirms that the organization has intermittent prediction traffic and wants to avoid managing endpoint capacity. The solution should remain easy to govern. Which option is the strongest recommendation?

A Use a nightly job even when users are waiting.
B Use only a data warehouse.
C Provision a permanently oversized endpoint.
D Evaluate serverless inference. check_circle
lightbulb

Explanation

Serverless inference can suit intermittent traffic when the team wants reduced capacity management. Why the strongest distractor fails: A permanently oversized endpoint may create unnecessary cost.

3
Domain 1: Fundamentals of AI and ML

[Easy] A global wholesaler needs to predict the numeric amount of electricity demand for the next hour. The choice must address the business requirement directly. Which action should the team take next?

A Use classification.
B Use Amazon Lex.
C Use clustering.
D Use regression. check_circle
lightbulb

Explanation

Regression predicts a continuous numeric value. Why the strongest distractor fails: Classification predicts categories rather than a numeric forecast.

4
Domain 1: Fundamentals of AI and ML

[Easy] During a design review, the AI program lead for an online retailer explains that the organization needs to identify sentiment and entities in customer reviews. The organization needs an approach that can be evaluated with measurable outcomes. Which recommendation is the best fit?

A Use Amazon Rekognition.
B Use Amazon Transcribe.
C Use Amazon Comprehend. check_circle
D Use Amazon Polly.
lightbulb

Explanation

Amazon Comprehend performs NLP tasks such as sentiment and entity detection. Why the strongest distractor fails: Amazon Rekognition analyzes images and video rather than review text.

5
Domain 1: Fundamentals of AI and ML

[Medium] During a design review, the application owner for a manufacturing group explains that the organization must minimize false positives when flagging legitimate transactions for manual review. The design must minimize unnecessary operational overhead. Which answer best addresses the requirement?

A Prioritize precision. check_circle
B Prioritize recall only.
C Use ROUG
E D. Use token count as the primary metric.
lightbulb

Explanation

Precision measures how many predicted positives are actually positive, which matters when false positives are costly. Why the strongest distractor fails: Recall emphasizes finding positives and can tolerate more false positives.

6
Domain 1: Fundamentals of AI and ML

[Medium] A data analyst is advising an automotive supplier. The organization needs an AI system that can reason about a goal, use approved tools, and complete a multi-step workflow. The team needs a production-ready recommendation. Which option is the strongest recommendation?

A Use an agentic AI approach. check_circle
B Use a relational database index only.
C Use text-to-speech synthesis only.
D Use a static classification threshold only.
lightbulb

Explanation

Agentic AI is relevant when the system must plan, use tools, and orchestrate steps toward a goal. Why the strongest distractor fails: A classifier can predict a label but does not orchestrate a multi-step workflow.

7
Domain 1: Fundamentals of AI and ML

[Medium] For a regulated internal workflow, a property-management company has support tickets labeled with the correct routing team and wants to learn a routing rule from examples. The reviewers want a clear reason for the recommendation. Which action should the team take next?

A Use unsupervised learning.
B Use reinforcement learning.
C Use supervised learning. check_circle
D Use text-to-speech.
lightbulb

Explanation

Supervised learning uses labeled examples to predict a known target. Why the strongest distractor fails: Unsupervised learning is used when target labels are absent.

8
Domain 1: Fundamentals of AI and ML

[Hard] An insurance company must make a regulated credit-risk prediction and needs a narrowly scoped, explainable output. The reviewers want a clear reason for the recommendation. The reviewers want the answer that best handles the stated tradeoff rather than a merely plausible alternative. Which recommendation is the best fit? Select the option containing the TWO best actions.

A 1. Evaluate a traditional ML model before choosing a foundation model. 2. Compare capabilities, latency, cost, and compliance constraints. check_circle
B 1. Always use the largest foundation model. 2. Skip explainability because the workload is regulated.
C 1. Evaluate a traditional ML model before choosing a foundation model. 2. Always use the largest foundation model.
D 1. Use an image-generation model. 2. Compare capabilities, latency, cost, and compliance constraints.
lightbulb

Explanation

Option A contains the complementary actions. Traditional ML can be a better fit for a narrow predictive problem with explainability constraints. The paired distractors include an unsuitable or incomplete action. A foundation model is not automatically the best choice for every AI use case.

9
Domain 1: Fundamentals of AI and ML

[Hard] For a multi-region business process, a university needs to extract printed text, forms, and table structure from scanned invoices. The recommendation must control cost without sacrificing the stated requirement. The reviewers want the answer that best handles the stated tradeoff rather than a merely plausible alternative. Which answer best addresses the requirement?

A Use Amazon Rekognition only.
B Use Amazon Textract. check_circle
C Use Amazon Personalize.
D Use Amazon Transcribe.
lightbulb

Explanation

Amazon Textract extracts text, forms, and tables from documents. Why the strongest distractor fails: Amazon Rekognition is not the specialized document-extraction service.

10
Domain 1: Fundamentals of AI and ML

[Hard] A food-delivery platform is evaluating whether an AI assistant creates business value after deployment. The recommendation must control cost without sacrificing the stated requirement. The reviewers want the answer that best handles the stated tradeoff rather than a merely plausible alternative. Which option is the strongest recommendation?

A Track business metrics such as ROI, customer feedback, and cost per user. check_circle
B Track only the number of subnets.
C Track only the number of IAM groups.
D Track only model file size.
lightbulb

Explanation

Business metrics show whether the workload achieves its intended outcome. Why the strongest distractor fails: Infrastructure counts do not establish business value.

Get all 1100 questions

Full access includes all questions, detailed explanations, PDF downloads, and timed mock exams.

quiz Start Free Practice download Free PDF