AWS Certified AI Practitioner AIF-C01 Practice Test 8
Question 70
An insurance provider receives recordings of approved claims interviews and needs searchable text transcripts. Before approving a rollout, an architect suggests Use Amazon Rekognition labels to extract spoken words from audio. Which mismatch should be documented for this precise design?
- Automatic speech recognition converts audio speech into written transcript content
- Rekognition image/video analysis is not a general speech transcription service
- Translation alone does not perform the requested speech recognition
- Polly generates audio from text rather than transcribing speech
Explanation
Correct Answer
B:This is the actual failure mode of the proposed approach: Use Amazon Rekognition labels to extract spoken words from audio. Rekognition image/video analysis is not a general speech transcription service
Incorrect Answers
A:Automatic speech recognition converts audio speech into written transcript content This explains the recommended design, but it is not the requested diagnosis of Use Amazon Rekognition labels to extract spoken words from audio.
C:This limitation concerns Use Amazon Translate without first obtaining a transcript; it does not diagnose the proposed Use Amazon Rekognition labels to extract spoken words from audio approach.
D:This limitation concerns Use Amazon Polly for the audio-to-text conversion; it does not diagnose the proposed Use Amazon Rekognition labels to extract spoken words from audio approach.
Question 71
An operations team must process a million stored images overnight rather than respond to user uploads instantly. Before approving a rollout, an architect suggests Replace image inference with an audio conversion service. Which mismatch should be documented for this precise design?
- Interactive calls add unnecessary per-request pressure when offline processing is allowed
- Bulk offline prediction need not maintain low per-request interactive latency
- Audio conversion does not produce requested visual predictions
- Re-training per input is not normal inference and adds avoidable cost
Explanation
Correct Answer
C:This is the actual failure mode of the proposed approach: Replace image inference with an audio conversion service. Audio conversion does not produce requested visual predictions
Incorrect Answers
A:This limitation concerns Choose only synchronous low-latency requests for each image; it does not diagnose the proposed Replace image inference with an audio conversion service approach.
B:Bulk offline prediction need not maintain low per-request interactive latency This explains the recommended design, but it is not the requested diagnosis of Replace image inference with an audio conversion service.
D:This limitation concerns Train a new model for each uploaded image; it does not diagnose the proposed Replace image inference with an audio conversion service approach.
Question 72
A model training project fails because currency fields contain strings and missing values are inconsistent. After selecting the recommended approach, which evidence MOST DIRECTLY demonstrates that the actual required outcome has been achieved?
- A component test invokes this alternate approach successfully: Skip training-data validation and rely only on more expensive compute
- A reviewed design record approves this competing approach: Assume deployment monitoring replaces initial data preparation
- A system inventory lists this competing approach as active: Tune output temperature on an existing generative model to repair a broken CSV
- Data validation reports consistent numeric fields and documented missing-value handling before training
Explanation
Correct Answer
D:Data validation reports consistent numeric fields and documented missing-value handling before training. This checks the business result rather than just a configured feature.
Incorrect Answers
A:Extra compute cannot guarantee consistent features or valid numeric inputs A component invocation does not validate the necessary end-to-end outcome.
B:Monitoring after deployment does not clean malformed training records Design approval is not evidence that the alternative meets the actual requirement.
C:Sampling temperature does not repair tabular training data quality An active setting cannot independently prove the requested business outcome.
Question 73
A binary disease-screening model misses many true positive cases although its overall accuracy looks high. During a design review, someone proposes Use mean absolute error as the primary binary confusion-matrix statistic. Which technical objection is specifically relevant to this proposal?
- Recall measures recovered actual positives and reveals the misses that accuracy can obscure
- Accuracy can appear high when negative cases dominate
- A continuous regression metric does not directly show missed positive classifications
- Inference spending does not count clinical positives missed
Explanation
Correct Answer
C:This is the actual failure mode of the proposed approach: Use mean absolute error as the primary binary confusion-matrix statistic. A continuous regression metric does not directly show missed positive classifications
Incorrect Answers
A:Recall measures recovered actual positives and reveals the misses that accuracy can obscure This explains the recommended design, but it is not the requested diagnosis of Use mean absolute error as the primary binary confusion-matrix statistic.
B:This limitation concerns Use only accuracy without inspecting class prevalence; it does not diagnose the proposed Use mean absolute error as the primary binary confusion-matrix statistic approach.
D:This limitation concerns Use token output cost as the medical false-negative measure; it does not diagnose the proposed Use mean absolute error as the primary binary confusion-matrix statistic approach.
Question 74
A simulated warehouse robot learns which route to choose from delayed rewards for delivery time and collisions. Which option BEST meets the business requirement?
- Use supervised OCR to read every shelf label as the entire control policy
- Use standard regression to output a single shipping price
- Use unsupervised clustering to assign static product categories
- Use reinforcement learning to optimize the reward-guided policy
Explanation
Correct Answer
D:The agent learns actions from rewards and environmental feedback across episodes The specified scenario makes this the most appropriate approach.
Incorrect Answers
A:Recognizing labels is not a learned navigation policy driven by rewards This prevents the approach from meeting the stated requirement.
B:A price estimate does not determine route actions over time This prevents the approach from meeting the stated requirement.
C:Clustering does not optimize sequences of robot actions against rewards This prevents the approach from meeting the stated requirement.
Question 75
An analytics team needs a managed way to experiment with and deploy its own specialized forecasting ML model. Which option BEST meets the business requirement?
- Treat an Amazon Bedrock managed FM invocation as a custom regression-training job
- Use Amazon SageMaker AI for the custom ML model lifecycle
- Use Amazon Polly to train the demand forecasting model
- Use Amazon Translate as the experiment tracking and model hosting service
Explanation
Correct Answer
B:SageMaker AI supports developing, training, evaluating and hosting custom ML workloads The specified scenario makes this the most appropriate approach.
Incorrect Answers
A:FM inference alone does not train the specified bespoke forecasting pipeline This prevents the approach from meeting the stated requirement.
C:Speech synthesis is not a general ML training and deployment platform This prevents the approach from meeting the stated requirement.
D:Translation does not supply a custom ML development lifecycle This prevents the approach from meeting the stated requirement.
Question 76
A streaming company has no audience-category labels but wants to discover patterns in viewing behavior. Before approving a rollout, an architect suggests Use translation of descriptions as the segmentation algorithm. Which mismatch should be documented for this precise design?
- Clustering can discover similar users without predefined class labels
- A supervised classifier requires an appropriate labeled target
- Language translation does not group viewing patterns by behavior
- Forecasting aggregates volume rather than discovering user groups
Explanation
Correct Answer
C:This is the actual failure mode of the proposed approach: Use translation of descriptions as the segmentation algorithm. Language translation does not group viewing patterns by behavior
Incorrect Answers
A:Clustering can discover similar users without predefined class labels This explains the recommended design, but it is not the requested diagnosis of Use translation of descriptions as the segmentation algorithm.
B:This limitation concerns Train supervised classification using nonexistent segment labels; it does not diagnose the proposed Use translation of descriptions as the segmentation algorithm approach.
D:This limitation concerns Use time-series forecasting to produce only next week's total traffic; it does not diagnose the proposed Use translation of descriptions as the segmentation algorithm approach.
Question 77
A manufacturer wants to classify defective and acceptable products using photographs of the assembly line. After selecting the recommended approach, which evidence MOST DIRECTLY demonstrates that the actual required outcome has been achieved?
- A reviewed design record approves this competing approach: Use speech synthesis to report labels without classifying images
- A system inventory lists this competing approach as active: Use speech-to-text transcription to interpret conveyor photographs
- Compare predictions with human-labeled holdout inspection photos
- A component test invokes this alternate approach successfully: Use text sentiment analysis of machine serial numbers
Explanation
Correct Answer
C:Compare predictions with human-labeled holdout inspection photos. This checks the business result rather than just a configured feature.
Incorrect Answers
A:Audio output does not identify the defect in the incoming image Design approval is not evidence that the alternative meets the actual requirement.
B:Transcription converts audio speech and cannot classify pixels An active setting cannot independently prove the requested business outcome.
D:Sentiment on identifiers does not inspect visual product defects A component invocation does not validate the necessary end-to-end outcome.
Question 78
A manufacturer wants to classify defective and acceptable products using photographs of the assembly line. Before approving a rollout, an architect suggests Use speech synthesis to report labels without classifying images. Which mismatch should be documented for this precise design?
- Audio output does not identify the defect in the incoming image
- Sentiment on identifiers does not inspect visual product defects
- The decisive signal is visual image content rather than plain-text or speech
- Transcription converts audio speech and cannot classify pixels
Explanation
Correct Answer
A:This is the actual failure mode of the proposed approach: Use speech synthesis to report labels without classifying images. Audio output does not identify the defect in the incoming image
Incorrect Answers
B:This limitation concerns Use text sentiment analysis of machine serial numbers; it does not diagnose the proposed Use speech synthesis to report labels without classifying images approach.
C:The decisive signal is visual image content rather than plain-text or speech This explains the recommended design, but it is not the requested diagnosis of Use speech synthesis to report labels without classifying images.
D:This limitation concerns Use speech-to-text transcription to interpret conveyor photographs; it does not diagnose the proposed Use speech synthesis to report labels without classifying images approach.
Question 79
An email service has thousands of messages labeled spam or legitimate and needs to classify future messages. Before approving a rollout, an architect suggests Use reinforcement learning with rewards for warehouse robots. Which mismatch should be documented for this precise design?
- A reward-driven control policy does not directly exploit the provided labeled email classes
- Regression predicts numerical quantities rather than the specified discrete spam class
- Clustering groups by similarity without directly learning the recorded spam labels
- Labels identify the target classes and allow the classifier to learn a decision boundary
Explanation
Correct Answer
A:This is the actual failure mode of the proposed approach: Use reinforcement learning with rewards for warehouse robots. A reward-driven control policy does not directly exploit the provided labeled email classes
Incorrect Answers
B:This limitation concerns Use regression to predict a continuous dollar amount; it does not diagnose the proposed Use reinforcement learning with rewards for warehouse robots approach.
C:This limitation concerns Use unsupervised clustering without consulting existing labels; it does not diagnose the proposed Use reinforcement learning with rewards for warehouse robots approach.
D:Labels identify the target classes and allow the classifier to learn a decision boundary This explains the recommended design, but it is not the requested diagnosis of Use reinforcement learning with rewards for warehouse robots.
Question 80
A hotel wants a conversational bot that collects check-in date and room type through voice or text turns. During a design review, someone proposes Use Amazon Translate alone to track hotel reservation state. Which technical objection is specifically relevant to this proposal?
- Lex is designed for conversational intent recognition and slot elicitation
- Image analysis is not a spoken reservation dialog manager
- Translation does not orchestrate multi-turn reservation fields
- Polly speaks text but does not manage user intents or slot collection
Explanation
Correct Answer
C:This is the actual failure mode of the proposed approach: Use Amazon Translate alone to track hotel reservation state. Translation does not orchestrate multi-turn reservation fields
Incorrect Answers
A:Lex is designed for conversational intent recognition and slot elicitation This explains the recommended design, but it is not the requested diagnosis of Use Amazon Translate alone to track hotel reservation state.
B:This limitation concerns Use Amazon Rekognition to infer date and room type from speech; it does not diagnose the proposed Use Amazon Translate alone to track hotel reservation state approach.
D:This limitation concerns Use Amazon Polly alone as the complete booking dialog engine; it does not diagnose the proposed Use Amazon Translate alone to track hotel reservation state approach.
