AIF-C01 Practice Exam 56 — 20 Questions

Question 1

A healthcare startup is building a patient-education chatbot. They want the model to answer only using an approved set of clinical guideline documents, and they must be able to update those documents weekly without retraining any model. They plan to use Amazon Bedrock. Which approach best meets these requirements?

Choose an answer to begin — your selection is scored in the full session.

Accepted Answer

Use Retrieval Augmented Generation by storing the guidelines in a vector store and retrieving relevant passages at inference time.. Retrieval Augmented Generation separates knowledge from model weights. The approved guidelines live in a vector store that can be updated weekly, and relevant passages are retrieved and passed to the model as context. This grounds answers in approved content and avoids the cost and delay of fine-tuning for every content change.

Answer

Fine-tune a foundation model in Amazon Bedrock each week on the updated guideline documents.

Answer

Increase the model's context window by raising the maxTokens parameter for every request.

Answer

Create a separate Bedrock agent for each clinical guideline document and route questions by keyword.

Question 2

A financial services firm wants its generative AI assistant to answer questions using its private policy documents, which change frequently. The firm does not want to retrain the foundation model each time a document is updated. Which approach best meets these requirements?

Accepted Answer

Use retrieval-augmented generation to fetch relevant passages from a knowledge base and include them in the prompt.. Retrieval-augmented generation decouples knowledge from model weights by retrieving relevant passages at query time and placing them in the prompt. Policy updates then require only re-indexing the documents, avoiding retraining entirely, while answers remain grounded in current private content. Temperature changes and larger context windows do not supply the missing knowledge on their own.

Answer

Expand the model's context window by requesting a larger variant of the same model.

Answer

Fine-tune the foundation model on the policy documents after every update.

Answer

Increase the model's temperature setting so it explores more of its training knowledge.

Question 3

A retail bank trained a loan-approval model using Amazon SageMaker. Before deployment, the compliance team asks the ML engineer to produce a report that shows, for each input feature, how strongly its values influence the model's predictions, so reviewers can confirm the model is not making decisions based on a protected attribute such as postal code. Which SageMaker Clarify capability should the engineer use to generate this feature-attribution report?

Accepted Answer

SHAP (Shapley Additive exPlanations) analysis via SageMaker Clarify. Feature attribution is needed to show which inputs drive predictions. SageMaker Clarify's SHAP analysis produces per-feature contribution values that reviewers can inspect to confirm the loan model is not leaning on a protected or proxy attribute. Drift detection, dataset-level bias metrics, and Trusted Advisor do not attribute model behavior to individual input features.

Answer

AWS Trusted Advisor security and fault-tolerance checks

Answer

Amazon SageMaker Model Monitor data drift detection

Answer

Amazon SageMaker Clarify pre-training bias metrics

Question 4

A media company uses Amazon Rekognition to automatically moderate user-uploaded images on its platform. The moderation team reports that some images containing nudity are being approved, while harmless images of sculptures are being rejected. The company wants to review the specific labels and confidence scores that Rekognition assigned to each image before deciding whether to appeal. Which action should the team take to obtain this information?

Accepted Answer

Enable Amazon Rekognition content moderation and inspect the ModerationLabels returned in the DetectModerationLabels response. DetectModerationLabels returns a ModerationLabels list where each entry includes a category name and a confidence score, giving the moderation team the exact evidence behind each decision. This supports appeals and threshold tuning. Custom Labels would replace the classifier, A2I adds human review rather than exposing existing scores, and CloudTrail logs API activity without capturing label content.

Answer

Enable AWS CloudTrail data events on the Amazon Rekognition API to capture label details

Answer

Use Amazon Augmented AI (A2I) to route every image to human reviewers before moderation

Answer

Configure Amazon Rekognition Custom Labels to train a new moderation model on the rejected images

Question 5

A media company is evaluating foundation models for an application that will summarize long earnings-call transcripts in English for internal analysts. The transcripts are up to 40,000 words, and the summaries must remain faithful to the source. Which TWO model characteristics are most important to evaluate for this workload? (Choose two.)

Accepted Answer

The model's quality on text summarization and faithfulness to source content. Long transcripts demand a model whose context window can accommodate the input, and financial summarization demands strong faithfulness so the output does not distort the source. Those two characteristics together determine whether the application can produce trustworthy summaries without excessive chunking. Image generation, speech quality, and edge deployment do not address the stated requirements.

Answer

The model's ability to run on edge devices without network connectivity

Answer

The model's text-to-speech voice quality

Answer

The model's ability to generate images from text

Question 6

A data science team is comparing two approaches for customizing a foundation model in Amazon Bedrock for a domain-specific classification task. Approach one is providing a small set of labeled examples directly in the prompt for each request. Approach two is fine-tuning the model on a larger labeled dataset. The team wants the lowest operational overhead and the fastest way to start, and their label set is small and changes frequently. Which statement best describes the trade-off they should consider?

Accepted Answer

Prompt-based examples offer lower operational overhead and adapt quickly to changing labels, while fine-tuning can improve consistency when a larger stable dataset is available.. The team's constraints are low operational overhead, fast start, and a small label set that changes often. Few-shot prompting embeds labeled examples in each request, needs no training job, and updates instantly. Fine-tuning is more appropriate for larger, stable datasets where consistent behavior is worth the training and maintenance effort, so it is not the best fit here.

Answer

Prompt-based examples require a dedicated training job in Amazon Bedrock, so they carry the same operational overhead as fine-tuning.

Answer

Fine-tuning eliminates the need for any labeled data because the model learns the task from unlabeled inputs automatically.

Answer

Fine-tuning is preferred because it always produces higher accuracy than prompt-based examples regardless of dataset size.

Question 7

A solutions architect must choose a foundation model for an application that summarizes lengthy internal audit reports. The reports average 60,000 tokens, and the summaries must reflect details from the beginning, middle, and end of each document. Cost per request matters, but recall of details is the top priority. Which model characteristic should drive the selection?

Accepted Answer

The model's maximum context window length. When a single request must contain an entire long document, the maximum context window is the hard constraint that decides feasibility. Parameter count, output modality, and latency influence quality or experience but do not determine whether 60,000 tokens can be processed without losing the details the task requires.

Answer

The model's inference latency percentile

Answer

The number of parameters in the model

Answer

The model's supported output modalities

Question 8

A product team is comparing foundation models for a customer-facing FAQ bot. They need low latency for real-time chat and want to pay only for what they use without managing servers. Which combination of model characteristic and AWS consumption model best fits these requirements?

Accepted Answer

A smaller, lower-latency model invoked on demand so billing is based on input and output tokens processed.. Real-time chat favors models that emit tokens quickly, and a smaller model usually wins on latency. On-demand invocation bills by tokens processed, so a variable-traffic FAQ bot pays only for actual usage and the provider handles scaling and infrastructure. Provisioned capacity and self-hosted GPUs suit steady high volume but add cost and operational burden here.

Answer

A large model with the highest parameter count, deployed on a provisioned throughput commitment billed hourly.

Answer

An open-weight model downloaded and hosted on AWS Lambda for per-request execution.

Answer

A multimodal model with image understanding, called through a self-managed Amazon EC2 GPU instance.

Question 9

A financial services firm is deploying a generative AI assistant that answers employee questions about internal policy documents. The security team requires that answers be traceable to source text and that the model not invent policy details. Which TWO techniques should be implemented to ground responses and reduce fabricated content? (Choose two.)

Accepted Answer

Instruct the model to answer only from the supplied context and to state when the context is insufficient. Grounding comes from combining retrieval of authoritative passages with instructions that confine the model to that evidence and permit abstention. Together they make answers traceable and suppress invention. Higher temperature, larger parameter counts, and removal of system instructions either increase variability, fail to supply evidence, or discard the constraints that keep output faithful.

Answer

Increase the temperature setting so the model explores a wider range of responses

Answer

Remove all system instructions so the model can respond more naturally to each question

Answer

Expand the model's parameter count by switching to the largest available variant

Question 10

A media company generates AI-written summaries of news articles using Amazon Bedrock and publishes them automatically. Legal counsel is concerned that the model might reproduce long verbatim passages from copyrighted source articles. The team wants a configurable safeguard that detects and filters responses containing text closely matching the source documents before publication. Which approach should they implement?

Accepted Answer

Implement a plagiarism or similarity detection step that compares generated text against the source corpus before publishing. Copyright overlap is a text-similarity problem, so the safeguard must compare generated summaries against the source article corpus and flag passages exceeding a similarity threshold. PII filters, word filters, and policy-based automated reasoning checks each address different risk categories and cannot measure verbatim reproduction of ordinary prose from a reference document.

Answer

Configure a sensitive information filter in Amazon Bedrock Guardrails to block personally identifiable information

Answer

Apply automated reasoning checks in Amazon Bedrock Guardrails to validate the summary against policy rules

Answer

Use a word filter in Amazon Bedrock Guardrails configured with the publisher's restricted terms

Question 11

A healthcare analytics team is evaluating whether a generative AI solution is appropriate for summarizing patient intake notes into structured clinical fields. They are concerned about the model producing confident but incorrect medical details. Which TWO practices best reduce the risk of fabricated content in this scenario? (Choose two.)

Accepted Answer

Instruct the model to respond with a specific phrase such as 'not stated' when the source note does not contain a requested field.. Fabrication drops when the model is anchored to the actual source text and is given permission to report that information is absent. Supplying the intake note in the prompt keeps generation evidence-based, and defining an explicit abstention phrase prevents the model from inventing values for missing fields. Randomness and self-reported confidence do not reliably reduce hallucination.

Answer

Ask the model to state its confidence as a percentage after each extracted field and trust fields scored above ninety percent.

Answer

Fine-tune the model on the entire historical patient record database to expand its clinical knowledge.

Answer

Raise the temperature to its maximum so the model produces diverse clinical phrasing across repeated runs.

Question 12

A logistics company wants to use machine learning to predict delivery times. A data scientist is preparing the project and must identify which characteristics describe supervised learning rather than unsupervised learning. (Choose two.)

Accepted Answer

The goal is to learn a mapping from input features to a known output so the model can predict on new data.. Supervised learning is defined by training on labeled examples and learning a mapping from inputs to a known output so the model can predict on new data. In the delivery-time scenario, historical trips with actual durations provide those labels. The remaining characteristics describe unsupervised learning, which discovers structure or groups without predefined outputs or ground truth.

Answer

The training process requires no historical outcomes and relies only on feature similarity.

Answer

The algorithm discovers hidden structure in data without any predefined output values.

Answer

The model is evaluated primarily by how well it separates data into groups with no ground truth.

Question 13

A government agency is building an AI assistant using Amazon Bedrock to help citizens understand eligibility rules for public benefits. The agency's legal team requires that the assistant never provide medical diagnoses, never discuss competitors' services, and refuse requests to draft legal documents. The agency also needs to log blocked interactions for periodic review. Which Amazon Bedrock feature should the team configure to enforce these restrictions and capture denials?

Accepted Answer

Amazon Bedrock Guardrails with denied topics and content filters. Bedrock Guardrails is the runtime content governance feature that enforces denied topics and content filters, blocking disallowed prompts and responses while producing logs for review. Evaluation jobs compare models offline, Provisioned Throughput manages capacity, and Agents extend capabilities through Lambda actions. Only Guardrails applies policy at inference time and captures the blocked interactions the agency must audit.

Answer

Amazon Bedrock Agents with action groups invoking Lambda functions

Answer

Amazon Bedrock Provisioned Throughput for dedicated model capacity

Answer

Amazon Bedrock model evaluation jobs comparing foundation models

Question 14

An insurance company uses an Amazon SageMaker model to set premium discounts. An internal audit finds that the model's error rates are substantially higher for one demographic group than another, even though overall accuracy is strong. The data science team must investigate and quantify this disparity before the model can be re-approved. Which approach should they take first?

Accepted Answer

Run a SageMaker Clarify bias analysis that computes subgroup performance metrics such as accuracy difference and disparate impact across the demographic facets. Before any remediation, the team must quantify the disparity by computing fairness metrics broken out by demographic facet. SageMaker Clarify bias analysis produces exactly these subgroup metrics, such as accuracy difference and disparate impact, giving the audit the evidence it needs. Retraining, monitoring, and added model capacity all act without first measuring the gap, so they cannot satisfy the re-approval requirement.

Answer

Increase the model's complexity by adding more layers to improve fit on the training data

Answer

Retrain the model on a larger dataset and compare overall accuracy before and after

Answer

Deploy the model with Amazon SageMaker Model Monitor to track data drift in production

Question 15

A hospital uses an Amazon SageMaker model to predict sepsis risk from electronic health records and displays a risk score to clinicians. An internal review finds that the model was trained on data from a single urban hospital and performs worse for patients from rural clinics. The review board asks the data science team to quantify and document this performance gap across patient subgroups before the model is expanded. Which approach should the team take?

Accepted Answer

Run SageMaker Clarify bias detection with the rural or urban clinic attribute as the sensitive facet and review the disparity metrics. SageMaker Clarify bias detection measures disparity across a chosen sensitive facet, so using clinic type as the facet quantifies the urban-versus-rural performance gap. The generated report documents the disparity for the review board. Learning rate tuning, Model Monitor drift detection, and Shadow Tests address optimization, operational drift, and traffic comparison respectively, none of which produce subgroup fairness metrics.

Answer

Retrain the model with a larger learning rate to improve overall accuracy on the combined dataset

Answer

Enable SageMaker Shadow Tests to compare the new model against the current clinical workflow

Answer

Deploy the model with SageMaker Model Monitor to detect data drift after expansion to rural clinics

Question 16

A product team is designing an internal tool that helps engineers understand a large legacy codebase. They want the assistant to explain functions, suggest refactors, and answer questions about dependencies. The team is comparing a general-purpose foundation model with a model pre-trained specifically on source code. Which consideration most strongly favors the code-specialized model?

Accepted Answer

Its pre-training corpus emphasized programming languages and code structure, improving performance on code comprehension tasks. Domain specialization during pre-training is what gives a code-focused model an edge on comprehension, refactoring, and dependency reasoning, because its training data emphasized programming constructs. Context window size, determinism, and knowledge of private repositories are independent of the training domain and are not improved simply by specializing on code.

Answer

It is guaranteed to have a larger context window than any general-purpose model

Answer

It will always produce deterministic output for identical prompts

Answer

It eliminates the need for retrieval or context when answering questions about private repositories

Question 17

A small insurance firm is selecting an AI service to classify customer emails by topic. The compliance officer insists that the provider publish clear documentation on how the model was built, its intended use, and its limitations, and that the firm be able to see model version details. Which AWS resource should the firm consult to evaluate these transparency characteristics of Amazon Bedrock foundation models?

Accepted Answer

Amazon Bedrock model cards in the model catalog. Transparency about how a model was built, its intended use, and its limits is documented in model cards. Amazon Bedrock publishes model cards in its model catalog, giving the insurer the version and suitability details the compliance officer wants. Compliance certifications, service health, and log queries describe infrastructure, availability, or runtime events, not model characteristics.

Answer

Amazon CloudWatch Logs insights queries

Answer

AWS Service Health Dashboard

Answer

AWS Artifact reports

Question 18

A retail bank is preparing to launch an Amazon Bedrock-based assistant that recommends credit card products to customers. The responsible AI review board requires the team to document how the system's outputs can be explained to regulators and customers. Which TWO actions best support explainability for this deployment? (Choose two.)

Accepted Answer

Store the retrieved source documents and the model's cited references alongside each generated recommendation for later review. Explainability for a regulated recommendation assistant requires preserving the evidence behind each output and communicating the automated nature of the decision to affected customers. Retaining retrieved sources and citations makes reasoning reproducible during audits, while AI disclosure with a human-review path gives customers transparency and contestability. Logging and catalog changes do not expose reasoning.

Answer

Increase the model's temperature setting so recommendations show more variety across customers

Answer

Enable verbose AWS CloudTrail logging of every InvokeModel API call for the assistant

Answer

Reduce the number of credit card products in the catalog so the model has fewer options to choose from

Question 19

A city government uses an Amazon SageMaker model to score affordable-housing applications. To meet its responsible AI commitments, the IT team must ensure that every automated decision can be traced back to the exact model version and training dataset used, and that changes are reviewed before deployment. Which combination of AWS practices best provides this accountability?

Accepted Answer

Register model versions in the SageMaker Model Registry with approval status and lineage, and require manual approval before deployment. Accountability for automated decisions requires a durable record of what produced each outcome and a controlled release process. The SageMaker Model Registry captures versioned model packages with approval status and lineage to training data, and its approval workflow enforces review before deployment. Scaling, instance sizing, and encryption improve performance or confidentiality but do not deliver traceability or approval gating.

Answer

Configure a SageMaker endpoint with a larger instance type to reduce inference latency for applicants

Answer

Store application data in an Amazon S3 bucket encrypted with AWS KMS customer managed keys

Answer

Enable automatic scaling on the SageMaker endpoint so it can handle fluctuating application volumes

Question 20

A fintech company wants its Amazon Bedrock assistant to answer customer questions only from its approved policy documents and to avoid fabricating answers when the documents do not cover a topic. The team plans to use a knowledge base with retrieval augmented generation. Which Bedrock Guardrails feature should they configure to detect and block responses that are not supported by the retrieved source passages?

Accepted Answer

A contextual grounding check with a grounding threshold and relevance threshold. Preventing fabricated answers requires checking whether a response is entailed by the retrieved context. Bedrock Guardrails contextual grounding performs exactly that comparison, and its grounding and relevance thresholds determine how strictly unsupported or off-topic responses are blocked. Harmful-content, PII, and word filters address toxicity or confidentiality but never verify factual support from source passages.

Answer

A sensitive-information filter that blocks personally identifiable information

Answer

A word filter that blocks a custom list of profane terms

Answer

Content filters for hate, violence, and insult