Knowledge + Practice

AWS Certified AI Practitioner AIF-C01 (AIF-C01) — Questions 526–600

1000 questions total · 14pages · All types, answers revealed

Take a mock exam Exam hub

Page 8 of 14

526

MCQeasy

A developer is using Amazon Bedrock Agents to build a multi-step reasoning bot that can query a SQL database and summarize results. Which service should be integrated as the action group's Lambda function to execute SQL queries?

A.Amazon Aurora

B.Amazon DynamoDB

C.Amazon OpenSearch Serverless

D.Amazon S3

AnswerA

Aurora is a relational SQL database that can be queried by the Lambda function.

Why this answer

The Lambda function can connect to any database, but for SQL querying, it would typically call an RDS or Aurora database. The integration is done via the Lambda code.

Full explanation →

527

MCQhard

A large enterprise has multiple teams deploying ML models on AWS. To ensure governance and accountability, they need to enforce that all models pass a fairness review before production deployment. Which SageMaker feature should they use to implement this approval workflow?

A.SageMaker Studio

B.SageMaker Experiments

C.SageMaker Model Monitor

D.SageMaker Model Registry

AnswerD

Model Registry supports versioning, metadata, and approval statuses for governance.

Why this answer

SageMaker Model Registry manages model versions and supports approval statuses (e.g., Approved, Rejected). This integrates with CI/CD pipelines to enforce review gates. Experiments, Model Monitor, and Studio do not provide such workflow capabilities.

Full explanation →

528

Multi-Selectmedium

Which THREE are SageMaker built-in algorithms suitable for regression tasks?

Select 3 answers

A.Linear Learner

B.K-Means

C.PCA

D.DeepAR

E.XGBoost

AnswersA, D, E

Linear Learner supports regression.

Why this answer

Linear Learner is a SageMaker built-in algorithm that supports both regression and classification tasks. For regression, it models the target variable as a linear combination of input features, optimizing for metrics like mean squared error. It is suitable for regression because it directly outputs continuous values.

Exam trap

The AIF-C01 exam often tests the distinction between supervised and unsupervised algorithms, and the trap here is that candidates may confuse dimensionality reduction (PCA) or clustering (K-Means) with regression tasks, assuming any algorithm that processes numeric data can perform regression.

Full explanation →

529

MCQeasy

A data scientist needs to predict house prices based on features like square footage, number of bedrooms, and location. Which type of machine learning is most appropriate for this task?

A.Classification

B.Reinforcement learning

C.Regression

D.Clustering

AnswerC

Regression models a continuous target variable based on input features.

Why this answer

Predicting a continuous value (price) from input features is a regression problem, a type of supervised learning.

Full explanation →

530

MCQeasy

A data scientist wants to host a pre-trained model on Amazon SageMaker for real-time inference with minimal latency. Which approach should they use?

A.Run inference using AWS Lambda with the model packaged as a container

B.Use SageMaker batch transform

C.Create a SageMaker asynchronous inference endpoint

D.Deploy the model on a SageMaker real-time endpoint

AnswerD

Real-time endpoints are designed for low-latency, synchronous inference.

Why this answer

Option D is correct because SageMaker real-time endpoints are designed for low-latency, synchronous inference. They keep the model loaded and ready to respond to individual requests, making them ideal for real-time applications where minimal latency is critical.

Exam trap

The AIF-C01 exam often tests the distinction between synchronous (real-time) and asynchronous inference patterns, and the trap here is that candidates may confuse 'asynchronous inference' with 'real-time' because both can handle requests, but only real-time endpoints guarantee minimal latency for individual predictions.

How to eliminate wrong answers

Option A is wrong because AWS Lambda has a maximum execution timeout of 15 minutes and limited memory (up to 10 GB), making it unsuitable for hosting large pre-trained models that require persistent, low-latency inference. Option B is wrong because SageMaker batch transform is an asynchronous, offline process for processing large datasets in batches, not for real-time inference with minimal latency. Option C is wrong because SageMaker asynchronous inference endpoints are designed for requests with large payloads and longer processing times, where immediate response is not required; they introduce queuing and processing delays that are incompatible with minimal latency requirements.

Full explanation →

531

Multi-Selecthard

A company uses Amazon Bedrock Agents for customer support. The agent needs to perform multi-step reasoning: first identify the customer's account, then check order status, and finally provide a resolution. Which THREE components must be configured to enable this workflow? (Select THREE.)

Select 3 answers

A.Agent orchestration to plan and execute multi-step reasoning

B.A Bedrock Knowledge Base with customer data

C.A Lambda function for each action group to execute the API calls

D.A Bedrock Guardrail to block out-of-scope questions

E.An action group for each API call (account lookup, order status)

AnswersA, C, E

The agent's orchestration capability handles the step-by-step reasoning.

Why this answer

Bedrock Agents use action groups to define tools, Lambda functions to execute them, and the agent orchestration handles multi-step reasoning. Knowledge Bases, Guardrails, and Playground are not required for this specific workflow.

Full explanation →

532

MCQmedium

A company wants to build a customer service chatbot that answers questions about their internal policy documents. The documents are updated monthly, and the team cannot afford to retrain a model each time. Which approach is MOST appropriate?

A.Fine-tune a base LLM on the policy documents monthly

B.Use a larger foundation model with a longer context window and paste all documents into each prompt

C.Train a custom model from scratch on the policy documents each month

D.Use Retrieval-Augmented Generation (RAG) with the policy documents indexed in a vector store

AnswerD

RAG retrieves relevant document chunks at query time, ensuring the chatbot always answers from the latest uploaded documents without any model retraining.

Why this answer

RAG (Retrieval-Augmented Generation) allows the LLM to retrieve relevant document sections at inference time, so knowledge stays current without retraining. The other options either require expensive retraining for each update or lack document grounding.

Full explanation →

533

Multi-Selecthard

A data engineer is using Amazon SageMaker Data Wrangler to prepare tabular data for ML. Which THREE data transformations are natively supported? (Choose three.)

Select 3 answers

A.One-hot encoding for categorical features

B.Audio feature extraction

C.Text vectorization using TF-IDF

D.Custom Python code via Pandas or Spark

E.Image resizing and normalization

AnswersA, C, D

One-hot encoding is a built-in transform in Data Wrangler.

Why this answer

Option A is correct because Amazon SageMaker Data Wrangler includes built-in support for one-hot encoding as a native transformation for categorical features. This transformation automatically creates binary columns for each category, which is essential for preparing tabular data for machine learning models that require numerical input.

Exam trap

AWS often tests the distinction between natively supported transformations in SageMaker Data Wrangler versus those requiring external services or custom scripts, leading candidates to mistakenly select audio or image processing options that are not part of Data Wrangler's built-in capabilities.

Full explanation →

534

Multi-Selectmedium

Which AWS services can be used to build, train, and deploy custom machine learning models? (Choose two.)

Select 2 answers

A.Amazon Polly

B.Amazon Lex

C.AWS Deep Learning AMIs

D.Amazon Rekognition

E.Amazon SageMaker

AnswersC, E

Deep Learning AMIs provide a customizable environment for building and training models.

Why this answer

AWS Deep Learning AMIs (C) are pre-configured Amazon Machine Images that include popular deep learning frameworks (TensorFlow, PyTorch, MXNet) and GPU drivers, allowing you to build, train, and deploy custom ML models on EC2 instances. Amazon SageMaker (E) is a fully managed service that provides end-to-end capabilities for building, training, and deploying custom ML models at scale, with built-in algorithms, automatic model tuning, and one-click deployment.

Exam trap

The trap here is that candidates confuse pre-built AI services (Polly, Lex, Rekognition) with platforms that allow custom model development, leading them to select services that only consume pre-trained models rather than build and train custom ones.

Full explanation →

535

MCQeasy

A media company uses Amazon Transcribe for automatic speech recognition. They discover the model has higher error rates for non-native English speakers. Which Responsible AI principle are they failing to uphold?

A.Fairness

B.Explainability

C.Robustness

D.Privacy

AnswerA

Fairness addresses disparate impact.

Why this answer

Option C is correct: Fairness requires equitable performance across groups. Option A is wrong: Explainability is about transparency. Option B is wrong: Privacy is about data protection.

Option D is wrong: Robustness is about reliability under varied conditions.

Full explanation →

536

MCQmedium

A company is using Amazon Bedrock to generate marketing copy. They want to evaluate the quality of the generated text. Which metric is MOST suitable for assessing the relevance and coherence of the content?

A.Accuracy

B.ROUGE-N

C.Perplexity

D.BLEU score

AnswerB

ROUGE-N compares n-gram overlap, suitable for summarization and copy.

Why this answer

ROUGE-N (Recall-Oriented Understudy for Gisting Evaluation) measures the overlap of n-grams between generated text and reference text, making it suitable for assessing relevance and coherence in content generation tasks like marketing copy. It evaluates how well the generated text captures key phrases and maintains logical flow, which aligns with the need to assess content quality beyond simple factual accuracy.

Exam trap

AWS often tests the distinction between metrics designed for translation (BLEU) versus summarization/generation (ROUGE), leading candidates to mistakenly choose BLEU for coherence evaluation when ROUGE is the correct choice for recall-based content assessment.

How to eliminate wrong answers

Option A is wrong because Accuracy is a classification metric (e.g., correct predictions/total predictions) and does not measure text relevance or coherence; it is irrelevant for generative text evaluation. Option C is wrong because Perplexity measures how well a language model predicts a sequence (lower is better for fluency) but does not directly assess relevance or coherence against a reference; it is a model-internal metric, not a quality metric for generated content. Option D is wrong because BLEU score (Bilingual Evaluation Understudy) is primarily designed for machine translation, focusing on precision of n-gram matches, and is less sensitive to recall and coherence in single-language text generation tasks like marketing copy.

Full explanation →

537

Multi-Selecthard

Which TWO practices help ensure responsible AI when deploying generative AI applications? (Select TWO.)

Select 2 answers

A.Deploy the model without any content filters to maximize creativity

B.Increase model size to improve accuracy at the expense of interpretability

C.Use only synthetic data for training to avoid privacy issues

D.Implement guardrails to filter harmful or inappropriate content

E.Monitor the model's outputs for bias and drift over time

AnswersD, E

Guardrails like Amazon Bedrock Guardrails help enforce content policies and prevent harmful outputs.

Why this answer

Implementing guardrails (e.g., content filtering) and monitoring for bias are key responsible AI practices. Using diverse training data is important but not a deployment practice. Publicly deploying without safeguards is irresponsible.

Full explanation →

538

MCQhard

A company uses Amazon Bedrock to generate code. They want to ensure the code follows security best practices and does not contain vulnerabilities. Which approach is most effective?

A.Implement a post-processing step using AWS WAF.

B.Use Amazon CodeGuru Security to review generated code.

C.Train a custom model on the company’s secure code.

D.Use a foundation model trained only on secure code.

AnswerB

CodeGuru Security automatically scans code for vulnerabilities and provides actionable recommendations.

Why this answer

Amazon CodeGuru Security reviews code for security vulnerabilities and provides recommendations. Using a model trained on secure code may not be sufficient; WAF is for web traffic; training a custom model requires significant effort and may not catch all issues.

Full explanation →

539

Multi-Selecthard

An organization is implementing governance for machine learning models using SageMaker. They need to track model versions, monitor for drift after deployment, and automatically decommission models that have been deprecated for over 30 days. Which THREE services or features should they use?

Select 3 answers

A.Amazon SageMaker Ground Truth

B.Amazon SageMaker Clarify

C.Amazon SageMaker Model Registry

D.Amazon SageMaker Projects (MLOps templates)

E.Amazon SageMaker Model Monitor

AnswersC, D, E

Model Registry manages model versions and lifecycle stages.

Why this answer

SageMaker Model Registry tracks versions, Model Monitor detects drift, and SageMaker Projects with MLOps can automate decommissioning. Clarify is for bias, not lifecycle management. Ground Truth is for labeling.

Full explanation →

540

Multi-Selecthard

A team is using Amazon Comprehend to analyze customer feedback for sentiment. They want to detect and mitigate potential bias against certain demographic groups. Which TWO approaches should they consider? (Choose TWO.)

Select 2 answers

A.Use AWS WAF to filter out biased comments.

B.Use AWS CloudTrail to audit API calls.

C.Use Amazon Rekognition to verify images.

D.Use SageMaker Clarify to compute bias metrics on the training data.

E.Use Comprehend custom classification with balanced training data across groups.

AnswersD, E

Clarify supports NLP bias detection and can analyze text datasets.

Why this answer

SageMaker Clarify can compute bias metrics on text data, and training with balanced data reduces bias. WAF is for web security, Rekognition is for image/video, CloudTrail is for auditing API calls – none are relevant to bias in NLP models.

Full explanation →

541

Multi-Selectmedium

A healthcare startup is deploying an AI system to assist in diagnosing skin conditions from images. They want to follow the NIST AI Risk Management Framework. Which THREE practices should they implement?

Select 3 answers

A.Use Amazon SageMaker to continuously monitor model performance and retrain as needed

B.Document the model's intended use, performance, and limitations in a model card

C.Establish a human-in-the-loop process for uncertain diagnoses

D.Archive all training data in Amazon S3 Glacier for long-term retention

E.Deploy the model on AWS Lambda for serverless inference

AnswersA, B, C

Continuous monitoring is part of risk management to detect and address issues over time.

Why this answer

The NIST AI RMF emphasizes governance, transparency, and risk management. Documenting model cards, establishing human oversight, and continuous monitoring align with these principles.

Full explanation →

542

Multi-Selectmedium

Which TWO actions are best practices when deploying foundation models on Amazon SageMaker for production? (Choose TWO.)

Select 2 answers

A.Manually warm up endpoints by sending dummy requests before traffic spikes.

B.Create a separate endpoint for each model to isolate traffic.

C.Use multi-model endpoints (MMEs) to serve multiple models on a single instance.

D.Implement inference pipelines to handle preprocessing and postprocessing steps separately.

E.Deploy models directly to production without load testing to avoid delays.

AnswersC, D

MMEs optimize resource utilization and reduce costs for multiple models.

Why this answer

Option C is correct because Amazon SageMaker Multi-Model Endpoints (MMEs) allow you to host multiple models on a single instance, which reduces hosting costs by sharing resources across models while still providing low-latency inference. This is a best practice for production deployments where you need to serve many models efficiently without provisioning separate endpoints for each.

Exam trap

AWS often tests the misconception that manual endpoint warm-up is necessary for production traffic spikes, but SageMaker's auto-scaling and built-in health checks handle this automatically, making option A a common distractor.

Full explanation →

543

Multi-Selecthard

A company is developing an AI system for resume screening. They want to ensure fairness and reduce bias. Which THREE steps should they take in accordance with the NIST AI Risk Management Framework and AWS responsible AI principles?

Select 3 answers

A.Establish a governance process for regular auditing and human review of decisions

B.Measure bias metrics across demographic groups using SageMaker Clarify

C.Optimize the model solely for overall accuracy

D.Ensure the training data includes diverse representation across demographic groups

E.Remove all sensitive attributes (e.g., gender, race) from the dataset

AnswersA, B, D

Governance and human oversight are critical for responsible AI, as emphasized in the NIST framework.

Why this answer

The NIST AI Risk Management Framework emphasizes measuring bias, ensuring diverse data, and establishing governance. Removing sensitive features alone is insufficient due to proxy correlations. Relying solely on accuracy ignores fairness.

The three correct steps cover measurement, data diversity, and governance.

Full explanation →

544

Multi-Selecthard

A developer is using the Amazon Bedrock InvokeModel API with streaming enabled. They want to process partial results as they arrive. Which THREE steps are necessary to implement streaming correctly? (Select THREE.)

Answer options not yet available.

Why this answer

To implement streaming, you must enable streaming in the invocation, parse the response stream, and handle each chunk (e.g., by appending to a buffer).

Full explanation →

545

MCQmedium

Refer to the exhibit. A user invokes Claude v2 using the AWS CLI. The response is truncated. What is the most likely cause?

A.The AWS CLI is missing the --endpoint-url parameter.

B.The max_tokens_to_sample is too low.

C.The model does not support this use case.

D.The prompt includes a stop sequence 'Assistant:'.

AnswerD

Claude uses 'Assistant:' as a stop sequence, causing it to stop generating after its response.

Why this answer

Option D is correct because the prompt includes the stop sequence 'Assistant:', which causes the model to halt generation as soon as it encounters that token sequence. In Claude v2, stop sequences are used to control the output length and structure; when the model generates the exact stop sequence, it truncates the response at that point, even if more content could have been produced.

Exam trap

The AIF-C01 exam often tests the distinction between token limits and stop sequences, where candidates mistakenly attribute truncation to max_tokens_to_sample when the actual cause is a configured stop sequence in the prompt or API parameters.

How to eliminate wrong answers

Option A is wrong because the --endpoint-url parameter is used to specify a custom endpoint for the AWS CLI, but its absence does not cause response truncation; it would instead result in a connection error or default endpoint usage. Option B is wrong because max_tokens_to_sample controls the maximum number of tokens the model can generate, but if it were too low, the response would be cut off at that token limit, not at a specific stop sequence; the question states the response is truncated, not that it reached a token limit. Option C is wrong because Claude v2 supports a wide range of use cases including text generation, and the model's capability is not the cause of truncation; truncation is explicitly controlled by stop sequences or token limits.

Full explanation →

546

MCQmedium

A healthcare company needs to build a GenAI application that summarizes patient discharge notes. The compliance team requires that no medical record identifiers (MRIs) appear in the model's output. Which Bedrock Guardrails feature should be configured?

A.PII detection and redaction

B.Topic denial

C.Grounding check

D.Content filtering

AnswerA

PII detection identifies and redacts personal/medical identifiers from the model's output.

Why this answer

PII detection in Guardrails can redact sensitive identifiers. Content filtering handles toxicity, topic denial blocks subjects, and grounding checks factual support — only PII detection specifically deals with identifiers.

Full explanation →

547

MCQeasy

A company uses Amazon Bedrock to generate product descriptions. They notice that the output sometimes contains incorrect information. What should they do to improve accuracy?

A.Increase the temperature parameter.

B.Implement Retrieval-Augmented Generation (RAG).

C.Use a larger foundation model.

D.Use AWS WAF to filter outputs.

AnswerB

RAG retrieves relevant information from a knowledge base to augment the prompt, improving factual accuracy.

Why this answer

Option B is correct because Retrieval-Augmented Generation (RAG) enhances the accuracy of foundation model outputs by grounding the generation in authoritative, up-to-date external knowledge sources. Instead of relying solely on the model's parametric memory, RAG retrieves relevant documents or data from a vector database (e.g., Amazon OpenSearch Serverless) and injects them into the prompt context, reducing hallucinations and incorrect information in product descriptions.

Exam trap

AWS often tests the misconception that simply using a larger or more powerful model (Option C) is the universal fix for accuracy issues, when in fact the root cause of hallucinations is often a lack of grounded, retrievable context that RAG specifically addresses.

How to eliminate wrong answers

Option A is wrong because increasing the temperature parameter makes the model's output more random and creative, which would likely increase, not decrease, the frequency of incorrect information. Option C is wrong because using a larger foundation model does not inherently fix factual accuracy; larger models can still hallucinate or produce outdated information without access to current or domain-specific data. Option D is wrong because AWS WAF is a web application firewall that filters HTTP traffic for security threats (e.g., SQL injection, XSS) and has no mechanism to validate or correct the factual accuracy of generated text.

Full explanation →

548

MCQmedium

A data scientist runs the above AWS CLI command and receives the error. What is the most likely cause?

A.The IAM role does not have permissions.

B.The model is being updated.

C.The model is being deprecated.

D.The model is not supported in the current AWS region.

AnswerD

Foundation models are region-specific; the chosen model may not be available in the region used.

Why this answer

The ModelNotReadyException typically indicates the model is not available in the current region. The model may not be supported or is still being deployed. The error does not suggest deprecation, updating, or permissions issues.

Full explanation →

549

MCQmedium

Refer to the exhibit. A company sets up a knowledge base for a customer support chatbot using Amazon Bedrock. Users report that the chatbot misses relevant details from long documents. Which change to the data source configuration would most likely improve retrieval?

A.Increase the chunk size in FIXED_SIZE chunking

B.Change chunking strategy to SEMANTIC

C.Add more documents to the S3 bucket

D.Change the embedding model to a larger one

AnswerB

Semantic chunking groups related content, preserving context and improving retrieval accuracy.

Why this answer

The chunking strategy is set to FIXED_SIZE, which may split documents into chunks that are too small or lose context. Switching to SEMANTIC chunking improves retrieval by grouping paragraphs with similar meaning.

Full explanation →

550

MCQeasy

A company wants to build a generative AI application that can summarize customer support tickets. They need to ensure the model stays up-to-date with the latest product documentation without retraining. Which AWS service would best support this requirement?

A.Amazon Bedrock with Retrieval Augmented Generation (RAG)

B.Amazon Comprehend

C.Amazon Rekognition

D.Amazon SageMaker Ground Truth

AnswerA

Amazon Bedrock supports RAG, which enables the model to retrieve current information from a knowledge base, keeping summaries up-to-date without retraining.

Why this answer

Option D is correct because Amazon Bedrock with RAG allows the model to retrieve and incorporate up-to-date information from external sources without retraining. Option A (Amazon Comprehend) is for NLP but not generative summarization with live updates. Option B (Amazon Rekognition) is for image/video analysis.

Option C (Amazon SageMaker Ground Truth) is for data labeling.

Full explanation →

551

MCQhard

A machine learning engineer is fine-tuning an Amazon Titan Text Premier model on a dataset of 50,000 legal contracts. The average token length per contract is 4,500 tokens. The model has a maximum context window of 8,000 tokens. What is the MOST efficient way to prepare the training data?

A.Truncate each contract to 3,800 tokens to leave room for input/output formatting

B.Chunk each contract into segments of 3,800 tokens and treat each segment as a separate training example

C.Use a model with a larger context window, such as Anthropic Claude 3 Opus

D.Pad each contract with special tokens to reach exactly 8,000 tokens

AnswerA

Truncating to a safe length below the limit ensures all samples fit without exceeding the context window, and no chunking is needed.

Why this answer

Truncating each contract to 3,800 tokens (leaving room for prompt tokens) ensures all samples fit within the context window and avoids the cost and complexity of chunking or padding.

Full explanation →

552

MCQeasy

A company uses Amazon Bedrock to power a generative chatbot for employee onboarding. Recently, some employees reported that the chatbot occasionally provides responses that contain biased or offensive language. The company has a strict policy for respectful communication. They want to implement a solution quickly without retraining the model. Which action should they take?

A.Add a human reviewer to approve every response.

B.Use a different foundational model known for unbiased outputs.

C.Enable Amazon Bedrock's built-in content moderation filters.

D.Fine-tune the model on a dataset of polite conversations.

AnswerC

Guardrails can be activated immediately to filter harmful content.

Why this answer

Option B is correct because Amazon Bedrock's built-in content moderation filters (Guardrails) can be applied immediately to filter biased or offensive content without retraining. Option A (fine-tuning) is time-consuming and requires a dataset. Option C (switch model) may not be quick and still could produce biased outputs.

Option D (human reviewer) is slow and not scalable.

Full explanation →

553

MCQeasy

A developer wants to experiment with a foundation model for code generation without writing any code. Which AWS service provides a playground for models like CodeWhisperer?

A.Amazon CodeGuru

B.AWS Lambda

C.Amazon SageMaker Studio

D.Amazon Bedrock Playground

AnswerD

Bedrock provides a no-code playground to test models like Claude or CodeWhisperer.

Why this answer

Amazon Bedrock Playground is the correct answer because it provides a web-based interface for experimenting with foundation models (FMs) from providers like AI21 Labs, Anthropic, Cohere, Meta, Stability AI, and Amazon itself, including the CodeWhisperer model for code generation. This allows the developer to test prompts, adjust parameters, and see model responses without writing any code, directly fulfilling the requirement of a no-code playground.

Exam trap

The trap here is that candidates may confuse Amazon SageMaker Studio (a full ML development environment) with a no-code playground, overlooking that Bedrock Playground is specifically designed for zero-code experimentation with foundation models like CodeWhisperer.

How to eliminate wrong answers

Option A is wrong because Amazon CodeGuru is a service for automated code reviews and application profiling, not a playground for experimenting with foundation models or code generation. Option B is wrong because AWS Lambda is a serverless compute service for running code in response to events, not a no-code environment for testing foundation models. Option C is wrong because Amazon SageMaker Studio is an integrated development environment (IDE) for building, training, and deploying machine learning models, which typically requires writing code (e.g., Python notebooks) and is not a simple playground for foundation model experimentation without coding.

Full explanation →

554

Multi-Selecthard

An AWS AI practitioner is designing a document processing pipeline using Amazon Textract and Amazon Comprehend. The pipeline must extract text from PDFs, detect entities, and classify documents into categories (e.g., invoice, contract, report). Which THREE steps should be included in the pipeline? (Choose three.)

Select 3 answers

A.Use Amazon Comprehend to train a custom classifier for document type

B.Use Amazon Personalize to recommend document categories

C.Use Amazon Textract to extract text from the PDF

D.Use Amazon Comprehend to detect entities such as dates and amounts

E.Use Amazon Rekognition to analyze images in the PDF

AnswersA, C, D

Comprehend allows training custom classifiers for document categorization.

Why this answer

The pipeline should first use Textract to extract text from PDFs. Then Comprehend can perform entity detection (e.g., dates, amounts). For document classification, a custom classifier built with Comprehend or SageMaker can categorize documents.

Rekognition is for images; Transcribe for audio; Polly for speech synthesis; Personalize for recommendations.

Full explanation →

555

MCQeasy

Which type of machine learning is used when a model learns to play a game by receiving rewards or penalties for its actions?

A.Reinforcement learning

B.Semi-supervised learning

C.Supervised learning

D.Unsupervised learning

AnswerA

Reinforcement learning uses a reward signal to learn optimal behavior through trial and error.

Why this answer

Reinforcement learning involves an agent learning to make decisions by interacting with an environment and receiving feedback in the form of rewards or penalties. Supervised learning uses labeled data, unsupervised learning finds patterns in unlabeled data, and semi-supervised learning uses a mix of labeled and unlabeled data.

Full explanation →

556

MCQmedium

A government agency is deploying an AI system to detect fraudulent benefit claims. The system uses a neural network trained on historical claims data, which includes a disproportionate number of false positives from a particular ethnic group due to historical over-policing. The agency must ensure the system does not perpetuate discrimination. They have a rigorous testing procedure but limited budget. The project lead wants to balance fairness with detection performance. Which combination of steps should they prioritize?

A.Remove the race feature from the model and rely on performance metrics alone

B.Replace the neural network with a logistic regression model retrained on the same data

C.Rebalance the training data to have equal representation across groups and evaluate using a fairness metric like equal opportunity

D.Use a post-hoc explainability tool to identify biased predictions and manually override them

AnswerC

Rebalancing data and using fairness metrics directly mitigate bias and measure progress.

Why this answer

The most effective approach is to rebalance the training data to be more representative and to use a fairness metric, such as equal opportunity, during evaluation. This directly addresses the data bias and quantifies fairness. Excluding race features may still leave proxies.

Using a simpler model may not eliminate bias if data is biased. Post-hoc explanations help understand bias but do not fix it.

Full explanation →

557

MCQhard

A company is deploying a generative AI model that produces text summaries of legal documents. To comply with responsible AI guidelines, which of the following is the most critical to ensure transparency?

A.Informing users that the summaries are generated by AI

B.Ensuring the model does not reflect biases from training data

C.Achieving high performance on summary quality metrics

D.Guaranteeing the summaries are factually accurate

AnswerA

Disclosing AI-generated content is a key transparency requirement.

Why this answer

Transparency requires that users know when they are interacting with AI. Disclosing that the summaries are AI-generated is fundamental. Accuracy, bias mitigation, and performance are important but do not directly address transparency about AI use.

Full explanation →

558

MCQmedium

A developer is building an image generation application using Stability AI's Stable Diffusion model on Amazon Bedrock. The application needs to generate high-resolution images (1024x1024) with consistent style across multiple prompts. Which approach should they use?

A.Use a higher number of inference steps

B.Set the same seed value for all invocations

C.Increase the CFG scale to 20

D.Invoke the model through the Converse API with a system prompt

AnswerB

A fixed seed makes the initial noise identical, leading to more consistent style and composition across generations.

Why this answer

Using a consistent seed value ensures that the random noise input is the same across generations, producing more consistent styles. Adjusting CFG scale or steps affects quality but not consistency, while invoking with different providers is unnecessary.

Full explanation →

559

Multi-Selectmedium

Which TWO techniques are commonly used to prevent overfitting in machine learning models? (Select TWO.)

Select 2 answers

A.Add more irrelevant features

B.Use cross-validation

C.Increase model complexity

D.Reduce the amount of training data

E.Use regularization

AnswersB, E

Cross-validation helps assess model generalization and can indicate overfitting.

Why this answer

Cross-validation helps prevent overfitting by partitioning the training data into multiple folds, training the model on different subsets, and validating on held-out portions. This provides a more robust estimate of model performance on unseen data and reduces the risk of memorizing noise in a single train-test split.

Exam trap

AWS often tests the misconception that adding more data or features always helps model performance, when in fact irrelevant features or reducing training data can worsen overfitting, and candidates may incorrectly associate 'more complexity' with better generalization.

Full explanation →

560

MCQmedium

A financial services firm fine-tuned a generative AI model on Amazon SageMaker to summarize quarterly reports. The summaries often miss key financial metrics such as revenue and profit margins. The fine-tuning dataset contained full reports with summaries that included these metrics. The model appears to understand the reports but omits critical numbers. Which course of action would most likely improve the summaries?

A.Re-fine-tune using a carefully crafted dataset that includes explicit instructions to include key metrics and provides examples of correct summaries

B.Increase the maximum number of tokens in the summary

C.Switch to a different pre-trained model like Claude instead of the current one

D.Implement a post-processing Lambda function that extracts metrics from the original report and appends them to the summary

AnswerA

Better alignment through example prompts and targets teaches the model to focus on essential numbers.

Why this answer

The fine-tuning dataset likely lacks explicit instruction in the prompts to include specific metrics. Re-fine-tuning with examples that emphasize extracting and reporting numbers, or using a format that forces structured output, would help. Increasing length may include more text but not guarantee key metrics.

Changing model or post-processing won't fix the underlying training deficiency.

Full explanation →

561

MCQeasy

A company builds an AI system that generates medical diagnoses. To ensure patient safety and allow oversight, the company wants a human to review all high-risk predictions before they are acted upon. Which AWS service should they use?

A.AWS Lambda

B.Amazon SageMaker Ground Truth

C.Amazon Augmented AI (A2I)

D.AWS Step Functions

AnswerC

A2I integrates with SageMaker or other ML services to route predictions for human review when confidence is low or other conditions are met.

Why this answer

Amazon Augmented AI (A2I) enables human review workflows for ML predictions, allowing you to define conditions that trigger human review. SageMaker Ground Truth is for data labeling, not review workflows.

Full explanation →

562

MCQhard

An ML engineer wants to store training data in a format optimized for linear data scanning and columnar access in SageMaker. Which format is most appropriate?

A.JSON

B.Image (JPEG/PNG)

C.Parquet

D.CSV

AnswerC

Parquet is columnar and optimized for analytical queries.

Why this answer

Parquet is a columnar storage format optimized for both linear data scanning and columnar access, making it ideal for training data in SageMaker. It reduces I/O by storing data by columns rather than rows, enabling efficient retrieval of specific features during model training.

Exam trap

AWS often tests the misconception that CSV is the most efficient format for training data, but Parquet's columnar storage and compression provide superior performance for linear scanning and columnar access in distributed ML pipelines.

How to eliminate wrong answers

Option A is wrong because JSON is a row-oriented text format that requires full parsing for columnar access, leading to high I/O overhead and slower linear scans. Option B is wrong because image formats like JPEG/PNG are binary and designed for visual data, not structured tabular data, and lack columnar access capabilities. Option D is wrong because CSV is a row-oriented text format that, while simple, requires scanning entire rows to access specific columns and lacks compression and schema optimization.

Full explanation →

563

MCQhard

A company uses Amazon SageMaker to host a model for fraud detection. The model must be re-evaluated for bias on a monthly basis. Which SageMaker feature can be used to detect bias in a trained model?

A.SageMaker Debugger

B.SageMaker Model Monitor

C.SageMaker Clarify

D.SageMaker Autopilot

AnswerC

Clarify provides bias detection and explainability.

Why this answer

SageMaker Clarify is the correct choice because it is specifically designed to detect bias in machine learning models and data. It provides built-in capabilities to evaluate bias metrics (e.g., difference in positive proportions, disparate impact) both before training (pre-training bias) and after training (post-training bias), making it suitable for the monthly re-evaluation requirement.

Exam trap

The trap here is that candidates confuse SageMaker Model Monitor (which monitors data drift) with bias detection, but Model Monitor does not evaluate model fairness or bias metrics.

How to eliminate wrong answers

Option A is wrong because SageMaker Debugger is used for monitoring training jobs in real time to detect issues like vanishing gradients or overfitting, not for bias detection. Option B is wrong because SageMaker Model Monitor focuses on detecting data drift and quality issues in deployed endpoints, not on evaluating model bias. Option D is wrong because SageMaker Autopilot automates the process of building, training, and tuning models, but it does not include built-in bias detection capabilities.

Full explanation →

564

MCQmedium

A company is using Amazon Bedrock's Converse API to build a conversational agent. They want the agent to maintain context across multiple turns. The agent should also be able to call external APIs to retrieve real-time data when needed. Which combination of features should they use?

A.Use the Converse API with streaming enabled and set max tokens to 4096

B.Use the InvokeModel API with a conversational prompt template and tool use configuration

C.Use the Converse API with system prompts and tool configuration

D.Use the InvokeModel API with a system prompt and store conversation history client-side

AnswerC

The Converse API supports system prompts, multi-turn conversation history, and tool use for API calls.

Why this answer

The Converse API supports multi-turn conversations via conversation history. Tool use allows the model to request external API calls. System prompts set the behavior.

Full explanation →

565

Multi-Selectmedium

A company is building a generative AI application to generate product descriptions from customer reviews. They want to use Amazon Bedrock to access a foundation model. Which TWO actions should the company take to ensure responsible AI practices?

Select 2 answers

A.Use a single foundation model without any customization to avoid bias.

B.Implement human review of all generated descriptions before publication.

C.Monitor and log model inputs and outputs for auditing.

D.Regularly evaluate model performance and fine-tune with diverse data.

E.Disable content filtering to allow maximum creativity.

AnswersB, C

Human review provides oversight to catch harmful or biased outputs.

Why this answer

Options A and C are correct. Implementing human review (A) ensures oversight and catches harmful outputs. Monitoring and logging (C) enables auditing and detection of misuse.

Option B is incorrect because using a single model does not automatically avoid bias; customization may be needed. Option D is incorrect because disabling content filtering increases risk of generating inappropriate content. Option E is plausible but not a requirement specific to responsible AI; evaluation is part of ongoing improvement but not the immediate action.

Full explanation →

566

MCQmedium

A team is using Amazon SageMaker to deploy a real-time inference endpoint. The endpoint must be accessible only from a specific IP range and must automatically scale based on request volume. Which configuration meets these requirements?

A.Deploy the endpoint in a public subnet with a security group allowing the IP range.

B.Configure the endpoint with a VPC and attach a security group that allows inbound traffic from the IP range, and enable automatic scaling for the endpoint.

C.Deploy the endpoint with a VPC and use a Network Load Balancer with target group health checks.

D.Deploy the endpoint with an AWS WAF ACL to filter by IP and enable auto scaling for the endpoint.

AnswerB

VPC security group restricts by IP and automatic scaling handles demand.

Why this answer

Option B is correct because Amazon SageMaker endpoints can be deployed within a VPC, allowing you to attach a security group that restricts inbound traffic to a specific IP range. Additionally, SageMaker supports automatic scaling for real-time endpoints using Application Auto Scaling, which adjusts the number of instances based on request volume metrics like InvocationsPerInstance.

Exam trap

The trap here is that candidates often confuse network-level access control (security groups in a VPC) with application-layer filtering (AWS WAF) or assume that a public subnet with a security group is sufficient, not realizing that SageMaker endpoints in a public subnet are still internet-facing and cannot be restricted to a specific IP range without a VPC.

How to eliminate wrong answers

Option A is wrong because deploying the endpoint in a public subnet exposes it to the internet, and a security group alone cannot restrict access to a specific IP range if the endpoint is publicly accessible; SageMaker endpoints in public subnets are not supported for IP-based restriction without a VPC. Option C is wrong because while a Network Load Balancer (NLB) can provide health checks and distribute traffic, SageMaker endpoints do not require an NLB for IP-based access control or scaling; the VPC and security group configuration already handles access control, and NLB is not a standard component for SageMaker endpoint deployment. Option D is wrong because AWS WAF is a web application firewall that operates at the application layer (HTTP/HTTPS) and is typically associated with API Gateway or CloudFront, not directly with SageMaker endpoints; SageMaker endpoints do not natively integrate with AWS WAF for IP filtering, and using WAF would not replace the need for VPC-based network controls.

Full explanation →

567

MCQhard

A team is fine-tuning an Amazon Titan Text model on a small dataset of legal documents. After fine-tuning, the model produces outputs that are factually incorrect and sometimes contradicts the training data. What is the most likely cause?

A.The batch size was too large, reducing gradient noise and causing mode collapse

B.The learning rate was set too high, causing the model to diverge

C.The model was fine-tuned for too many epochs on a small dataset, leading to overfitting

D.The base model was not designed for legal domain tasks

AnswerC

Overfitting on a small dataset causes the model to memorize examples and fail to generalize, producing erratic outputs.

Why this answer

Overfitting on a small dataset can cause the model to memorize noise and fail to generalize, leading to incorrect outputs.

Full explanation →

568

MCQmedium

Refer to the exhibit. A developer has attached this IAM policy to their user. When trying to invoke the Anthropic Claude v2 model using the Bedrock runtime, they receive an AccessDeniedException. Which change to the policy would resolve the issue?

A.Add the bedrock:InvokeModelWithResponseStream action

B.Change the Action to bedrock:ListFoundationModels

C.Change the Resource to arn:aws:bedrock:us-east-1::foundation-model/anthropic.claude-v2

D.Remove the Resource element and set Effect to Deny

AnswerC

Correct. This ARN matches the Claude v2 model, allowing invocation.

Why this answer

The policy grants access only to the Titan model resource. To invoke Claude v2, the resource must match the Claude model ARN. Adding other actions or removing the resource condition would not grant the correct access; listing models does not allow invocation.

Full explanation →

569

MCQhard

A healthcare company wants to use generative AI to automatically generate patient summary reports from electronic health records (EHRs). The solution must be HIPAA compliant and data must not leave AWS. They plan to use Amazon Bedrock with a foundation model. The EHR data is stored in Amazon S3 and contains protected health information (PHI). Which approach best meets compliance requirements?

A.Use Amazon Bedrock with a HIPAA-eligible account, enable encryption with KMS, and de-identify PHI in the prompt

B.Use a publicly available foundation model API outside AWS for better accuracy

C.Use Amazon Comprehend Medical for entity extraction and then feed results into a model on Amazon Bedrock without de-identification

D.Use Amazon SageMaker with a public model from the internet without encryption

AnswerA

Bedrock is HIPAA-eligible when used with AWS Organizations and BAA; de-identification and KMS encryption protect PHI.

Why this answer

Amazon Bedrock operates within a HIPAA-eligible environment when configured appropriately, and using AWS KMS for encryption and not storing PHI in prompts (using de-identification) can maintain compliance. Using public models or non-HIPAA services would violate requirements. SageMaker with encryption can also be HIPAA-eligible, but Bedrock with proper settings is simpler.

Full explanation →

570

MCQeasy

An ML team notices that the training accuracy is 99% but validation accuracy is only 72%. Which concept best describes this situation?

A.Cross-validation error

B.Overfitting

C.Bias-variance tradeoff

D.Underfitting

AnswerB

Large gap between high training accuracy and lower validation accuracy is classic overfitting.

Why this answer

Overfitting occurs when the model learns training data too well, including noise, and fails to generalize to new data. Underfitting would show low accuracy on both sets.

Full explanation →

571

MCQmedium

A company uses Amazon Bedrock Agents to automate order processing. The agent needs to call an internal database to check inventory. Which component should be used to integrate the database query?

A.Knowledge base

B.Action group

C.Guardrail

D.Prompt template

AnswerB

Correct. Action groups enable agents to invoke external systems via Lambda.

Why this answer

Bedrock Agents use action groups to define external tools or APIs. Each action group has an associated AWS Lambda function that performs the actual work (e.g., querying a database). The agent can call the action group via natural language instructions.

Full explanation →

572

MCQmedium

A machine learning engineer notices that a SageMaker training job failed due to insufficient permissions to access a KMS-encrypted S3 bucket. The training job's IAM role has S3 access permissions. What should be done to resolve the issue?

A.Create a new KMS key and re-encrypt the data

B.Assign the SageMakerFullAccess policy to the role

C.Add a kms:Decrypt permission to the IAM role for the specific KMS key used to encrypt the S3 bucket

D.Change the S3 bucket's default encryption to S3-managed keys (SSE-S3)

AnswerC

The role needs decryption permissions to access the encrypted data.

Why this answer

The training job failed because the IAM role lacks permission to decrypt the KMS key used for S3 server-side encryption. Even with S3 access permissions, SageMaker cannot read encrypted objects without the kms:Decrypt action on the specific KMS key. Adding kms:Decrypt to the role's policy for that key resolves the issue.

Exam trap

The trap here is that candidates assume S3 permissions alone are sufficient, overlooking that KMS-encrypted objects require explicit kms:Decrypt permissions on the IAM role, not just S3 bucket policies or managed policies like SageMakerFullAccess.

How to eliminate wrong answers

Option A is wrong because creating a new KMS key and re-encrypting the data is unnecessary and disruptive; the existing key can be used if the IAM role is granted the proper decrypt permission. Option B is wrong because SageMakerFullAccess is an AWS managed policy that does not include KMS permissions for customer-managed keys; it only grants basic SageMaker and S3 access, so it would not resolve the KMS decryption failure. Option D is wrong because changing the bucket's default encryption to SSE-S3 removes KMS encryption, which may violate security or compliance requirements, and is an overreaction when a simple IAM permission update can fix the issue.

Full explanation →

573

Multi-Selecthard

A healthcare startup is deploying an AI model to assist with diagnosis. They want to comply with the EU AI Act, which classifies medical AI as high-risk. Which THREE requirements must they fulfill? (Choose three.)

Select 3 answers

A.Provide technical documentation and logs for traceability

B.Establish a risk management system throughout the AI system's lifecycle

C.Ensure human oversight to prevent or minimize risks

D.Minimize the amount of training data to only necessary data

E.Allow the model to make final decisions without human review

AnswersA, B, C

High-risk AI systems must be transparent with documentation and logging for traceability.

Why this answer

The EU AI Act for high-risk systems requires risk management, human oversight, and transparency including documentation. Data minimization is a GDPR principle, not specific to the AI Act; automated decisions without human review are not allowed for high-risk.

Full explanation →

574

MCQeasy

Which of the following is a key advantage of using a pre-trained foundation model over training a model from scratch?

A.Reduces the amount of labeled data and compute resources needed

B.Eliminates the need for any fine-tuning

C.Guarantees perfect accuracy on domain-specific tasks

D.Allows the model to work offline without any cloud infrastructure

AnswerA

Transfer learning from a pre-trained model drastically reduces data and compute requirements.

Why this answer

Pre-trained models have already learned general language patterns on large corpora, requiring far less data and compute for downstream tasks.

Full explanation →

575

MCQmedium

A company wants to use a third-party foundation model from Amazon Bedrock but is concerned about data privacy because the model provider might store prompts and responses. How should they address this concern?

A.Enable Amazon Bedrock model invocation logging to capture all interactions

B.Review the third-party model provider's data handling policy and choose a model that does not retain data

C.Use Amazon Macie to scan prompts before they are sent

D.Use AWS KMS to encrypt the prompts and responses before sending to the model

AnswerB

The provider's policy governs data retention; choosing a compliant model is the best approach.

Why this answer

Each third-party model provider in Bedrock has its own data handling policies. Customers should review those policies and can choose models that do not store data or use Bedrock features like Guardrails to redact sensitive data. However, the direct action is to review the provider's policy and select a model that meets privacy requirements.

Full explanation →

576

MCQmedium

A company is using Amazon Bedrock to generate code snippets. They notice the model occasionally generates code that fails to compile. What is the most effective way to improve code quality without retraining?

A.Reduce the temperature parameter to 0 for deterministic output.

B.Increase the max token limit to allow the model to complete the code fully.

C.Fine-tune the model on a dataset of correct code snippets.

D.Use few-shot prompt engineering with correct code examples and formatting instructions.

AnswerD

Examples help the model understand the expected output and reduce errors.

Why this answer

Option B is correct because prompt engineering with examples and constraints can guide the model to produce more accurate code. Option A is wrong because reducing temperature increases determinism but doesn't guarantee correctness. Option C is wrong because fine-tuning is expensive and may overfit.

Option D is wrong because increasing max tokens may lead to more errors.

Full explanation →

577

Multi-Selecteasy

A data scientist needs to use Amazon SageMaker to train a model and must ensure that the training data and the model artifacts are encrypted using customer-managed KMS keys. Which TWO resources can be encrypted with KMS keys in this scenario? (Select TWO.)

Select 2 answers

A.SageMaker ML storage volume (attached to training instances)

B.SageMaker notebook instances

C.Amazon SageMaker Model Registry

D.Amazon CloudWatch Logs

E.S3 buckets storing training data

AnswersA, E

SageMaker encrypts the local storage volume using a KMS key.

Why this answer

SageMaker can use KMS keys to encrypt the ML storage volume (attached to training instances) and S3 buckets storing training data. Model artifacts in S3 can also be encrypted with KMS.

Full explanation →

578

MCQeasy

A developer wants to test different prompt variations for a chatbot without making repeated API calls. Which Amazon Bedrock feature can help compare model responses?

A.Model evaluation on Amazon SageMaker

B.Amazon Bedrock Playground

C.AWS Security Token Service (STS)

D.Amazon CloudWatch Logs

AnswerB

The playground allows developers to test and compare prompts interactively.

Why this answer

Option D is correct because Amazon Bedrock Playground provides an interactive interface to experiment with prompts and compare outputs side by side. Option A (CloudWatch Logs) is for monitoring. Option B (Model evaluation on SageMaker) is for offline evaluation.

Option C (AWS STS) is for security tokens.

Full explanation →

579

Multi-Selecthard

A financial services company is deploying a large language model (LLM) on Amazon Bedrock for customer-facing applications. The compliance team mandates that the model must not generate any content containing personally identifiable information (PII). Additionally, the company wants to ensure that the model only answers questions related to its product documentation and refuses off-topic queries. Which THREE Bedrock Guardrails configurations should be applied?

Select 3 answers

A.Denied topics

B.PII redaction

C.Word filters

D.Grounding check

E.Content filters

AnswersA, B, E

Define topics the model should not discuss, such as off-topic queries.

Why this answer

To prevent PII leakage, enable PII redaction. To restrict to product documentation, use topic restrictions (denied topics for off-topic). To prevent harmful content, use content filters.

Word filters are for specific words, and grounding check is for hallucination prevention but not directly for topic restriction.

Full explanation →

580

MCQeasy

In a binary classification problem, the model predicts majority class for all inputs. What is this issue called?

A.High bias

B.Overfitting

C.High variance

D.Underfitting

AnswerA

Predicting majority class for all inputs indicates the model has high bias and is underfitting.

Why this answer

When a model predicts the majority class for all inputs, it indicates that the model is too simplistic and fails to capture the underlying patterns in the data. This is a classic symptom of high bias, where the model makes strong assumptions about the data distribution, leading to systematic underperformance on the minority class. In machine learning, high bias often results from an overly simple algorithm or insufficient model capacity, causing the model to underfit the training data.

Exam trap

The AIF-C01 exam often tests the distinction between 'high bias' and 'underfitting' as separate concepts, where underfitting is the symptom and high bias is the cause, so candidates may incorrectly select underfitting when the question explicitly asks for the name of the issue.

How to eliminate wrong answers

Option B (Overfitting) is wrong because overfitting occurs when the model learns noise and details from the training data too well, resulting in high variance and poor generalization, not a constant prediction of the majority class. Option C (High variance) is wrong because high variance typically leads to models that are overly sensitive to small fluctuations in the training data, producing different predictions for similar inputs, not a uniform majority class output. Option D (Underfitting) is a related concept but is not the specific term for the issue described; underfitting refers to the model's inability to capture the training data's patterns, which can cause high bias, but the question asks for the name of the issue itself, which is high bias.

Full explanation →

581

Multi-Selecteasy

A developer is new to Amazon Bedrock and wants to understand the components of tokenization and context windows. Which TWO statements are correct? (Select TWO.)

Select 2 answers

A.A larger context window always produces better quality responses

B.Tokenization splits text only on whitespace

C.The context window determines the maximum number of tokens the model can process in a single request

D.The cost of a request is based on the number of tokens per second

E.Tokens can be words or subwords, depending on the tokenizer

AnswersC, E

Context window limits input plus output tokens.

Why this answer

Tokens are units of text (words or subwords); context window is the maximum number of tokens the model can process. Tokens can be words or subwords; context window is a limit, not a guarantee of quality; tokens per second measures throughput, not cost; tokenization is not based solely on whitespace.

Full explanation →

582

MCQhard

An organization uses AWS Lake Formation to govern access to data used for machine learning in Amazon SageMaker. They want to ensure that a particular IAM role used by SageMaker can only query a subset of columns in a table containing sensitive customer data. Which Lake Formation permission should be granted to the role?

A.Grant DESCRIBE on the table

B.Grant INSERT on the table

C.Create a view with only the allowed columns and grant SELECT on the view

D.Grant SELECT on the table with column-level filter

AnswerD

Lake Formation allows granting SELECT on specific columns, achieving column-level access control.

Why this answer

Lake Formation supports column-level permissions via Data Permissions. You can grant SELECT permission on specific columns of a table to an IAM role, restricting access to only those columns.

Full explanation →

583

MCQeasy

Which of the following is a key principle of responsible AI according to AWS?

A.Complexity

B.Speed

C.Profitability

D.Transparency

AnswerD

Transparency is one of the six responsible AI principles from AWS.

Why this answer

Option B is correct because AWS emphasizes transparency as a key principle of responsible AI, alongside fairness, accountability, privacy, and robustness. Profitability (A), speed (C), and complexity (D) are not principles.

Full explanation →

584

MCQmedium

A company uses Amazon SageMaker Ground Truth to label a dataset for a binary classifier. To reduce labeling bias, which workforce configuration is most appropriate?

A.Automatic labeling with Active Learning

B.Public workforce with no qualification

C.Private workforce of domain experts

D.Vendor managed workforce

AnswerC

Domain experts provide high-quality, consistent labels, reducing bias.

Why this answer

Option D is correct because using an AWS-managed workforce (Mechanical Turk) often involves multiple workers per item, but for sensitive tasks, a private workforce of domain experts ensures consistency and reduces bias. Public (A) may introduce uncontrolled bias. Vendor (B) is not a built-in type.

Automatic (C) is not a workforce type.

Full explanation →

585

MCQhard

A financial services company is building a generative AI application using Amazon Bedrock. They need to ensure the model does not generate responses that violate regulatory topics, such as specific prohibited financial advice. Which Bedrock feature should they use to block entire topics?

A.Topic denial

B.PII detection

C.Content filtering

D.Grounding check

AnswerA

Topic denial lets you define specific topics that the model must not discuss, such as prohibited financial advice.

Why this answer

Bedrock Guardrails' topic denial capability allows you to define denied topics and prevent the model from generating responses related to those topics.

Full explanation →

586

MCQmedium

A machine learning model achieves 99% accuracy on the training set but only 65% on the test set. Which phenomenon is the model experiencing?

A.Bias-variance tradeoff

B.Data leakage

C.Overfitting

D.Underfitting

AnswerC

Overfitting results in high training performance but poor test performance.

Why this answer

High training accuracy with significantly lower test accuracy indicates overfitting, where the model learns training data noise and fails to generalize.

Full explanation →

587

Multi-Selecthard

A company needs to analyze customer feedback from social media posts (text) and determine the sentiment (positive, negative, neutral). They also need to extract key phrases and entities mentioned in the posts. Which THREE AWS services can be combined to accomplish this? (Choose THREE.)

Select 3 answers

A.Amazon Transcribe

B.Amazon Rekognition

C.Amazon Translate

D.Amazon Comprehend

E.Amazon Personalize

AnswersA, C, D

Transcribe can convert audio/video social media posts to text for further analysis.

Why this answer

Amazon Comprehend provides sentiment analysis and key phrase/entity extraction. Alternatively, one could use Transcribe (speech-to-text) if the posts are audio/video, but the scenario mentions text posts. Translate may be needed if posts are in multiple languages.

Rekognition is for images, and Personalize is for recommendations. The most direct combination is Comprehend for NLP tasks, Translate for multilingual support, and Transcribe if any content is audio-based. However, the scenario says 'text' posts, so Transcribe may not be needed.

A better combination: Comprehend (sentiment+entities), Translate (if multilingual), and SageMaker (custom model) but that is overkill. The intended correct set is: Amazon Comprehend, Amazon Translate (to standardize language), and Amazon Transcribe (to convert any audio/video to text). But since the scenario specifies 'social media posts' which are usually text, a more appropriate set might be Comprehend, Transcribe (for video posts), and Translate.

I'll choose the combination that fits the scenario of processing text and extracting entities: Comprehend for NLP, Transcribe for any audio/video content, and Translate for multilingual support.

Full explanation →

588

MCQmedium

A company uses Amazon Comprehend to analyze customer sentiment. They discover the model performs poorly on text with slang from underrepresented groups. What is the most responsible action?

A.Restrict model use to only standard English

B.Remove slang from input before inference

C.Adjust the confidence threshold only for those groups

D.Collect more representative training data including slang

AnswerD

Representative data reduces bias.

Why this answer

Option D is correct because the core principle of responsible AI requires that models be trained on data that is representative of the populations they serve. Amazon Comprehend's sentiment analysis is a supervised machine learning model; its poor performance on slang from underrepresented groups indicates a training data bias. Collecting more representative training data, including that slang, directly addresses the root cause by enabling the model to learn the linguistic patterns of those groups, improving fairness and accuracy without restricting access or masking the problem.

Exam trap

The trap here is that candidates may choose a quick-fix technical workaround (like removing slang or adjusting thresholds) instead of recognizing that the responsible AI approach requires addressing the root cause of bias through data representativeness, which is a core ethical and technical principle tested in the AIF-C01 exam.

How to eliminate wrong answers

Option A is wrong because restricting model use to only standard English is a discriminatory practice that excludes underrepresented groups, violating responsible AI principles of fairness and inclusivity; it does not fix the model's bias but rather avoids it. Option B is wrong because removing slang from input before inference is a data preprocessing workaround that does not address the underlying model bias; it discards valuable linguistic data and can alter the true sentiment of the text, leading to inaccurate results. Option C is wrong because adjusting the confidence threshold only for those groups is a post-hoc tuning that does not correct the model's learned bias; it may reduce false positives but does not improve the model's understanding of slang, and it introduces inconsistent decision boundaries that can be seen as unfair.

Full explanation →

589

MCQmedium

A company wants to build a customer service chatbot that answers questions about their internal policy documents. The documents are updated monthly, and the team cannot afford to retrain a model each time. Which approach is MOST appropriate?

A.Fine-tune a base LLM on the policy documents monthly

B.Use Retrieval-Augmented Generation (RAG) with the policy documents indexed in a vector store

C.Use a larger foundation model with a longer context window and paste all documents into each prompt

D.Train a custom model from scratch on the policy documents each month

AnswerB

RAG retrieves relevant document chunks at query time, ensuring the chatbot always answers from the latest uploaded documents without any model retraining.

Why this answer

RAG (Retrieval-Augmented Generation) allows the LLM to retrieve relevant document sections at inference time, so knowledge stays current without retraining. The other options either require expensive retraining for each update or lack document grounding.

Full explanation →

590

MCQeasy

Which metric is most appropriate for evaluating a classification model when false positives are costly?

A.Precision

B.F1 score

C.Recall

D.Accuracy

AnswerA

Precision is the fraction of true positives among predicted positives, addressing false positives.

Why this answer

Precision is the most appropriate metric when false positives are costly because it measures the proportion of true positive predictions among all positive predictions (TP / (TP + FP)). A high precision indicates that when the model predicts a positive class, it is very likely correct, minimizing the number of false positives. This directly aligns with the business requirement to avoid costly false alarms.

Exam trap

The AIF-C01 exam often tests the distinction between precision and recall by framing a cost scenario, and the trap here is that candidates confuse 'costly false positives' with 'costly false negatives' and incorrectly choose recall or F1 score without analyzing which error type is being penalized.

How to eliminate wrong answers

Option B (F1 score) is wrong because it is the harmonic mean of precision and recall, balancing both false positives and false negatives; it does not specifically penalize false positives more heavily. Option C (Recall) is wrong because it measures the proportion of actual positives correctly identified (TP / (TP + FN)), which is useful when false negatives are costly, not false positives. Option D (Accuracy) is wrong because it considers overall correct predictions (TP + TN) divided by total predictions, which can be misleading in imbalanced datasets and does not isolate the cost of false positives.

Full explanation →

591

Multi-Selecteasy

A developer is building an application that uses Amazon Bedrock to answer questions based on a large internal knowledge base. The knowledge base contains PDFs, Word documents, and web pages. Which TWO AWS services are commonly used together to implement a Retrieval-Augmented Generation (RAG) architecture on AWS? (Select TWO.)

Select 2 answers

A.Amazon Bedrock Knowledge Bases

B.Amazon SageMaker Ground Truth

C.AWS Glue

D.Amazon Athena

E.Amazon OpenSearch Serverless

AnswersA, E

Knowledge Bases is the managed service for creating a RAG pipeline with document ingestion and retrieval.

Why this answer

Amazon Bedrock Knowledge Bases provides a managed RAG solution that can ingest documents from various sources. Amazon OpenSearch Serverless (or another vector database) stores the embeddings for retrieval.

Full explanation →

592

MCQmedium

A company wants to use Amazon Bedrock to translate customer emails from English to Spanish. The emails contain occasional personal names and addresses. Which Guardrail configuration should be applied to protect customer privacy?

A.Grounding check

B.PII detection

C.Topic denial

D.Content filtering

AnswerB

Correct. PII detection identifies and can mask personal data.

Why this answer

PII detection in Bedrock Guardrails can identify and redact personally identifiable information such as names and addresses. Content filtering, topic denial, and grounding check do not specifically target PII.

Full explanation →

593

MCQhard

Refer to the exhibit. A data scientist ran a training job on Amazon SageMaker and it failed. Which action should the data scientist take FIRST to resolve the issue?

A.Request a service limit increase for the instance type

B.Use a different AWS region

C.Enable spot training

D.Use a different instance type that is available in the region

AnswerD

The error clearly states the instance type is unsupported; switching to an available type resolves it.

Why this answer

Option D is correct because the error indicates that the requested instance type is not available in the current region due to capacity constraints. The first step is to switch to a different instance type that is available in the same region, as this is the quickest and most direct way to resolve the provisioning failure without requiring service limit increases or changing regions.

Exam trap

AWS often tests the distinction between capacity unavailability (which requires switching instance types) and service limits (which require a limit increase), leading candidates to mistakenly request a limit increase when the real issue is temporary capacity constraints.

How to eliminate wrong answers

Option A is wrong because a service limit increase addresses the maximum number of instances you can run, not the immediate unavailability of a specific instance type in the region. Option B is wrong because using a different AWS region is a more drastic step that may introduce latency, data residency issues, or additional costs; the first action should be to try an alternative instance type within the same region. Option C is wrong because enabling spot training does not resolve the unavailability of the instance type; spot instances still require available capacity for the requested instance type.

Full explanation →

594

MCQmedium

A financial services company needs to use Amazon Bedrock to generate customer-facing content that must comply with strict regulatory guidelines. The company wants to minimize the risk of the model generating non-compliant content. Which technique should the company implement?

A.Use a higher temperature setting to make the model more conservative

B.Reduce the context window to limit the amount of input the model can see

C.Fine-tune the model on a dataset of compliant content only

D.Implement a guardrail that denies prohibited topics and enforces compliance rules

AnswerD

Guardrails allow you to define policies that block or filter content matching specific criteria, directly addressing compliance.

Why this answer

A guardrail is a configurable policy in Amazon Bedrock that can filter content based on predefined rules, such as denying topics or content that violates regulatory guidelines. This provides a systematic way to enforce compliance.

Full explanation →

595

MCQeasy

A data scientist is prototyping a text summarization application using Amazon Bedrock. They want to quickly test different prompts and models without writing code. Which AWS service or feature should they use?

A.AWS Cloud9

B.Amazon SageMaker Studio

C.AWS Lambda

D.Bedrock Playground

AnswerD

Bedrock Playground provides a console-based chat interface to test models and prompts without any code.

Why this answer

Bedrock Playground is a no-code interface within the AWS Console for experimenting with models and prompts. It is designed for rapid prototyping without any coding.

Full explanation →

596

MCQmedium

A startup wants to generate high-quality images from text descriptions using Amazon Bedrock. They need to create realistic images of products for an e-commerce catalog. Which model provider should they choose?

A.Stability AI

B.Anthropic

C.Mistral AI

D.Cohere

AnswerA

Stability AI offers diffusion models that generate images from text prompts.

Why this answer

Stability AI provides diffusion models (e.g., Stable Diffusion) that are specialized for text-to-image generation on Bedrock.

Full explanation →

597

Multi-Selecteasy

Which TWO are benefits of using Amazon SageMaker JumpStart for foundation models? (Choose 2)

Select 2 answers

A.Built-in fine-tuning scripts and notebooks

B.No coding required to fine-tune models

C.Automatic scaling without any configuration

D.Pre-trained foundation models available in the catalog

E.Free unlimited usage for all models

AnswersA, D

JumpStart provides prebuilt notebooks and scripts for common fine-tuning tasks.

Why this answer

JumpStart provides pre-trained foundation models and built-in fine-tuning scripts, accelerating development. It does require some coding for customization. It offers many models but not unlimited free usage (charges apply for infrastructure).

Scaling is configurable but not fully automatic without setup.

Full explanation →

598

MCQmedium

Refer to the exhibit. A data scientist used SageMaker Clarify to evaluate bias in a binary classification model predicting loan approval. The exhibit shows bias metrics for the female facet. What does the analysis indicate about the model's impact on the female group?

A.The metrics are within acceptable thresholds, so no action is needed.

B.The model shows a high positive bias toward the female group.

C.The model has a post-training accuracy difference indicating a negative bias against the female group.

D.The model exhibits a pre-training class imbalance but no post-training bias.

AnswerC

Negative AD means lower accuracy for the female group.

Why this answer

The post-training Accuracy Difference (AD) of -0.22 indicates that the model's accuracy for the female group is 22% lower than for the male group, representing a negative bias. Pre-training metrics show some imbalance (CI=0.2) and DPL=-0.15 indicating underrepresentation and lower proportion of positive labels, but the post-training metric directly shows performance disparity. Option A is incorrect because AD is present.

Option C is incorrect because AD is negative. Option D is incorrect because the threshold typically is 0.1 for bias detection, and -0.22 exceeds it.

Full explanation →

599

MCQmedium

A company is using Amazon Bedrock to generate embeddings for a semantic search application. They want to ensure that semantically similar phrases (e.g., "car" and "vehicle") produce similar vector representations. Which type of model should they use?

A.An embedding model like Amazon Titan Embeddings

B.An image generation model like Stable Diffusion

C.A text generation model like Anthropic Claude

D.A multimodal model that supports both text and image input

AnswerA

Titan Embeddings generate high-quality vector representations optimized for semantic similarity.

Why this answer

Embedding models are specifically designed to produce dense vector representations that capture semantic meaning, enabling similarity search.

Full explanation →

600

Multi-Selecthard

A machine learning team is building a binary classifier using Amazon SageMaker. The dataset has 10,000 features and 1,000 samples. The model overfits severely. Which TWO approaches are MOST likely to reduce overfitting? (Choose two.)

Select 2 answers

A.Increase the batch size to the full dataset

B.Use a neural network with more layers

C.Perform feature selection to reduce the number of features

D.Add L2 regularization to the loss function

E.Train the model for more epochs

AnswersC, D

Reducing features reduces dimensionality and overfitting risk.

Why this answer

With high dimensionality and small sample size, overfitting is common. Adding L1 or L2 regularization (via a penalty in the loss function) can reduce model complexity. Reducing the number of features through feature selection or dimensionality reduction also helps.

Increasing model complexity (more layers) would worsen overfitting; training longer also worsens it; increasing batch size might help slightly but not as directly.

Full explanation →

Page 8 of 14

All pages

1 2 3 4 5 6 7 8 9 10 11 12 13 14

Practice AIF-C01 by domain

Target a specific domain to shore up weak areas.

Applications of Foundation Models Fundamentals of AI and ML Fundamentals of Generative AI Guidelines for Responsible AI Security, Compliance and Governance for AI Solutions

See all domains with question counts →

AWS Certified AI Practitioner AIF-C01 AIF-C01 Questions 526–600 | Page 8/14 | Courseiva