AWS AI Practitioner (AIF-C01) Practice Exam

Free sample questions

1. Sample · question 1 · Rekognition image content moderation

   Cobblestone Photo runs a photo-sharing app and wants to automatically detect explicit or violent images that users upload, using a pre-trained AWS service instead of building its own model. Which service should it use?

   • A.Amazon Textract
   • B.Amazon Rekognitioncorrect
   • C.Amazon Comprehend
   • D.Amazon Translate

   Why: Amazon Rekognition provides pre-trained image and video analysis, including content moderation that detects unsafe or inappropriate content such as explicit or violent imagery. Textract extracts text and data from documents, Comprehend analyzes text, and Translate converts text between languages.

   Open this question on its own page →
2. Sample · question 2 · Comprehend Medical clinical text extraction

   Hartfield Clinic wants to extract medical conditions, medications, and dosages from unstructured physicians' notes using a managed, pre-trained service. Which AWS service is designed for this?

   • A.Amazon Lex
   • B.Amazon Polly
   • C.Amazon Rekognition
   • D.Amazon Comprehend Medicalcorrect

   Why: Amazon Comprehend Medical is a natural language processing service that detects and extracts medical information, such as conditions, medications, dosages, and tests, from unstructured clinical text. Polly converts text to speech, Rekognition analyzes images and video, and Lex builds conversational interfaces.

   Open this question on its own page →
3. Sample · question 3 · Polly text-to-speech narration

   Marchbank News wants to offer an audio version of every written article, generated automatically in a natural-sounding voice. Which AWS service should it use?

   • A.Amazon Pollycorrect
   • B.Amazon Transcribe
   • C.Amazon Lex
   • D.Amazon Comprehend

   Why: Amazon Polly converts text into lifelike speech in many voices and languages, which suits automatic audio versions of articles. Transcribe does the opposite and converts speech to text, Lex builds chatbots and voice bots, and Comprehend analyzes the meaning of text.

   Open this question on its own page →
4. Sample · question 4 · Lex conversational chatbot intents

   Tidewater Utilities wants to build a chatbot for its website and phone line that recognizes what customers want, such as 'report an outage' or 'check my bill', and collects the details needed to complete each request. Which AWS service is designed for building this conversational interface?

   • A.Amazon Polly
   • B.Amazon Personalize
   • C.Amazon Lexcorrect
   • D.Amazon Textract

   Why: Amazon Lex builds conversational interfaces for voice and text, using intents to capture what the user wants and slots to collect the details needed to fulfil the request. Personalize generates recommendations, Textract extracts document data, and Polly only converts text to speech.

   Open this question on its own page →
5. Sample · question 5 · Transcribe and Translate call pipeline

   Mirabel Travel records customer calls in Spanish and wants English text versions of each call for its analysts. Which AWS services should it combine? (Select TWO.)

   • A.Amazon Textract
   • B.Amazon Transcribecorrect
   • C.Amazon Translatecorrect
   • D.Amazon Rekognition
   • E.Amazon Polly

   Why: Amazon Transcribe converts the Spanish audio into Spanish text, and Amazon Translate then translates that text into English. Polly generates speech from text, Textract extracts text from documents and images, and Rekognition analyzes images and video, so none of them handles these two steps.

   Open this question on its own page →
6. Sample · question 6 · Personalize real-time recommendations

   Bellhaven Books wants to show each website visitor product recommendations based on their browsing and purchase history, using a managed service that does not require ML expertise. Which AWS service fits?

   • A.Amazon Personalizecorrect
   • B.Amazon Comprehend
   • C.Amazon Polly
   • D.Amazon Textract

   Why: Amazon Personalize trains recommendation models from user interaction data, such as views and purchases, and serves personalized recommendations without requiring ML expertise. Comprehend analyzes text, Polly converts text to speech, and Textract extracts data from documents.

   Open this question on its own page →
7. Sample · question 7 · SageMaker Data Wrangler data preparation

   A data scientist at Orrin Mobility wants to import data from several sources, explore it visually, and apply transformations such as handling missing values and encoding categories before training, with minimal custom code. Which SageMaker AI capability is designed for this?

   • A.Amazon Augmented AI (Amazon A2I)
   • B.Amazon SageMaker Model Cards
   • C.Amazon Bedrock Agents
   • D.Amazon SageMaker Data Wranglercorrect

   Why: SageMaker Data Wrangler, available in SageMaker Canvas, lets users import, explore, visualize, and transform data for ML with built-in transformations and little code. Model Cards document models, Amazon A2I adds human review of predictions, and Bedrock Agents orchestrate multi-step generative AI tasks.

   Open this question on its own page →
8. Sample · question 8 · Deep learning learns features from raw data

   Engineers at Fallow Vision explain why they chose deep learning instead of a traditional ML algorithm to recognize objects in raw photos. Which reason is most accurate?

   • A.Deep learning models always produce fully interpretable decisions.
   • B.Multi-layer neural networks can learn useful features directly from raw pixels instead of relying on hand-crafted features.correct
   • C.Deep learning models need no training data.
   • D.Deep learning works only with small, structured tabular datasets.

   Why: Deep learning uses neural networks with many layers that learn increasingly abstract features, such as edges, shapes, and objects, directly from unstructured data like images, reducing the need for manual feature engineering. Deep learning typically needs large amounts of training data, is harder to interpret than simpler models, and is especially strong on unstructured data rather than limited to small tabular datasets.

   Open this question on its own page →
9. Sample · question 9 · Stop sequences end generation

   A developer at Greystone Apps wants a text generation model to stop producing output as soon as it writes the marker '###'. Which inference parameter should the developer set?

   • A.Temperature
   • B.Top-k
   • C.A stop sequencecorrect
   • D.The context window size

   Why: A stop sequence is a string that, when generated, causes the model to stop producing further tokens, which is useful for ending output at a known marker. Temperature and top-k shape which tokens are sampled, and the context window is a fixed property of the model that limits how much text it can process.

   Open this question on its own page →
10. Sample · question 10 · SageMaker Feature Store shared features

    Ormsby Manufacturing's data science teams keep re-creating the same input features, such as average machine temperature over the last hour, for different models, and the values used in training sometimes differ from those used at inference. They want a central repository to store, share, and reuse features for both training and low-latency real-time inference. Which SageMaker AI capability fits?

    • A.Amazon Polly
    • B.Amazon SageMaker Feature Storecorrect
    • C.Amazon SageMaker Model Cards
    • D.Amazon Rekognition

    Why: SageMaker Feature Store is a managed repository for ML features, with an online store for low-latency lookups during real-time inference and an offline store for training and batch use, so teams can share consistent feature definitions across models. Model Cards document models, Polly converts text to speech, and Rekognition analyzes images and video.

    Open this question on its own page →
11. Sample · question 11 · SageMaker Model Registry version catalog

    Hadley Freight's ML team wants a central catalog of model versions in which each version has an approval status, so that only approved versions are deployed by its CI/CD pipeline. Which SageMaker AI feature provides this?

    • A.Amazon SageMaker Ground Truth
    • B.Amazon Macie
    • C.Amazon Translate
    • D.Amazon SageMaker Model Registrycorrect

    Why: SageMaker Model Registry catalogs models in model groups, tracks versions and their metadata, and manages approval status so that deployment pipelines can promote only approved versions. Ground Truth labels data, Macie discovers sensitive data in Amazon S3, and Translate converts text between languages.

    Open this question on its own page →
12. Sample · question 12 · Test set for final unbiased evaluation

    A data scientist at Wilmot Analytics splits a dataset into training, validation, and test sets. What is the main purpose of the test set?

    • A.To provide a final, unbiased estimate of performance on unseen data after model selection and tuning are completecorrect
    • B.To tune hyperparameters repeatedly during training iterations
    • C.To supply additional examples for the model to learn from
    • D.To store production inference requests for later monitoring

    Why: The test set is held back until the end and used once to estimate how the chosen model will perform on new, unseen data. The validation set is used during development to tune hyperparameters and compare models, which is why it cannot give an unbiased final estimate. Training data is what the model learns from, and production requests are captured separately for monitoring.

    Open this question on its own page →