AWS Certified AI Practitioner Readiness Test

10 questions, balanced across the exam's topics. Answer each one, then submit for your practice readiness estimate and per-topic breakdown. This is a practice estimate — not a prediction or guarantee of your official exam result.

Question 1

A team assumes a foundation model will only state true facts. What is wrong with this assumption?

Models are always truthful Users never lie Training data is perfect The model cannot verify facts it was not given

Question 2

An instructor asks: which of the following can provide intelligent search across enterprise documents?

Amazon Transcribe Amazon Comprehend Amazon Polly Amazon Kendra

Question 3

A voice assistant fails for speakers with accents. What most directly addresses the fairness gap?

More parameters Longer training Higher temperature Representative training data

Question 4

Which of the following means: tailoring messages and offers to individual preferences and behavior

Content moderation Personalized marketing Demand forecasting Call center analytics

Question 5

A company wants to discover where PII lives in its data lake. Which service fits?

Manual review of every file Deleting all S3 data Macie finds PII in S3 Ignoring S3 buckets

Question 6

A neural network has an input layer, three hidden layers, and an output layer. What makes this deep learning rather than a shallow model?

The use of labeled data The multiple hidden layers The size of the training batch The choice of optimizer

Question 7

Which of the following means: tracking the origin, movement, and transformations of data used to train models

Data privacy Data lineage Transparency Explainability

Question 8

Why can an embedding model find relevant documents even when they share no keywords with the query?

Embeddings for the query and documents share a vector space Keywords must match exactly The model must be retrained per query Documents must be in the same file

Question 9

A RAG system sometimes gives wrong answers. What is the most common root cause?

Embeddings are perfect Retrieval failures cause most ungrounded answers The model is malicious Users are at fault

Question 10

An instructor asks: which of the following can extract text and structured data from scanned documents?

Amazon Textract Amazon Bedrock Amazon SageMaker Amazon Rekognition

Unanswered questions count as incorrect. Your answers are graded on the server; no answer key is sent to your browser.

← Back to the AWS Certified AI Practitioner page

About Contact Privacy Policy Terms of Service Refund Policy

Home Exams Simulations WorkTrials GameScope Account