AIF-C01 Practice Exam 57 — 20 Questions

Question 1

A marketing team uses Amazon Bedrock to generate promotional copy for a global campaign. A reviewer discovers that some outputs include biased stereotypes about certain nationalities. The team wants a configurable control that detects and blocks this category of harmful content before the copy reaches reviewers. Which Amazon Bedrock feature should they configure?

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

Accepted Answer

Amazon Bedrock Guardrails content filters for hate and insult categories. The requirement is an inline, configurable control that detects and blocks harmful stereotyped content at generation time. Amazon Bedrock Guardrails content filters provide exactly that, with tunable strength for hate and insult categories. Throughput reservation, latency alarms, and offline model evaluation jobs address capacity, operations, and model selection rather than runtime content safety.

Answer

Amazon Bedrock model evaluation jobs that compare foundation models on benchmark datasets

Answer

Amazon CloudWatch alarms on the model's invocation latency

Answer

Amazon Bedrock provisioned throughput to reserve dedicated model capacity

Question 2

A public sector agency is building a chatbot on Amazon Bedrock that answers citizen questions about benefits eligibility. The agency must ensure the chatbot never provides medical or legal advice, and that responses stay grounded in the agency's official policy documents rather than the foundation model's general knowledge. Which Amazon Bedrock Guardrails configuration should the team apply?

Accepted Answer

Configure a denied topics policy for medical and legal advice, and use contextual grounding checks with the official policy documents supplied as the grounding source.. Two constraints must be enforced: no medical or legal advice, and answers grounded in official policy documents. A denied topics policy names and blocks those advice categories at both input and output. Contextual grounding checks compare each response against the supplied policy documents and intervene when support is weak, which prevents the model from answering from its own general knowledge.

Answer

Configure a sensitive information filter to redact medical terms and legal citations, and enable the profanity filter at HIGH strength to keep responses professional.

Answer

Configure automated reasoning checks against a formal policy of eligibility rules, and rely on the foundation model's built-in knowledge to answer general questions.

Answer

Configure a word filter containing a blocklist of medical and legal terms, and set the guardrail action to NONE so responses are flagged for human review instead of blocked.

Question 3

A hospital's IT team wants a generative AI assistant that answers patient-billing questions using only the hospital's internal policy documents, without retraining the model. Which approach should they use?

Accepted Answer

Retrieval Augmented Generation (RAG) by querying a vector store of the policy documents and passing retrieved passages to the model. Retrieval Augmented Generation supplies the model with relevant, current policy text at query time, so responses are grounded in the hospital's own documents. Fine-tuning changes model weights and needs retraining, temperature alters sampling randomness, and a system prompt alone provides no source content. Only retrieval of the policy passages gives accurate, updatable answers without retraining.

Answer

Fine-tuning the foundation model on the full set of policy documents

Answer

Increasing the model's temperature setting so it explores more of its pretrained knowledge

Answer

Prepending a system instruction telling the model to answer only from hospital policy

Question 4

A media company's generative AI writing assistant produces fluent but sometimes fabricated statistics. The team wants to reduce these hallucinations without changing the foundation model. Which action best addresses the root cause?

Accepted Answer

Ground the model's responses by supplying verified reference documents in the prompt and instructing it to cite them. Fabricated statistics stem from the model generating text without an authoritative factual anchor. Supplying verified documents and requiring citations ties each claim to a source, which reduces invention while leaving the model untouched. Token limits, parameter count, and sampling parameters influence length, capacity, and randomness respectively, none of which ground output in verified facts.

Answer

Raise the top-p sampling value to make token selection more deterministic

Answer

Lower the maximum token limit for each response

Answer

Switch the model to a smaller parameter count to reduce creativity

Question 5

A financial services firm is evaluating foundation models for a loan-summary assistant. They must consider both model characteristics and operational constraints. Which TWO factors most directly affect whether a candidate model can be deployed to meet their requirements? (Choose two.)

Accepted Answer

The model's maximum context window length relative to the size of the loan documents. Deployability hinges on concrete constraints: whether the model can ingest the full loan document within its context window, and whether it is offered in the Region where data must remain. Parameter count, community popularity, and marketing presentation do not determine whether the model can satisfy the workload and compliance requirements.

Answer

The popularity of the model among hobbyist developers on public forums

Answer

The color scheme used in the model provider's marketing materials

Answer

The number of parameters reported in the model's architecture diagram

Question 6

A developer wants a foundation model to reliably return a structured record with fixed fields for downstream processing. The model currently returns free-form prose that breaks parsing. Which technique most directly improves output reliability for this use case?

Accepted Answer

Define a JSON schema and use constrained decoding or structured output features so the model can only emit tokens valid under that schema. Structured output reliability comes from constraining decoding to a schema, which makes invalid tokens impossible rather than merely unlikely. Prompt-based requests for JSON and retry loops reduce but do not eliminate malformed output, raising temperature increases variability, and shortening the prompt removes guidance without adding enforcement. Schema-constrained generation is the direct fix.

Answer

Increase the model's temperature so it generates more varied field values

Answer

Shorten the prompt so the model has fewer instructions to follow

Answer

Add the phrase 'respond in JSON' to the user prompt and retry on parse failures

Question 7

A solutions architect needs to build a generative AI application that can invoke foundation models from Amazon and third-party providers through a single, unified API without managing any infrastructure. The architect wants the fastest path to a working prototype using AWS-native tooling. Which AWS service should the architect choose?

Accepted Answer

Amazon Bedrock. Amazon Bedrock is purpose-built for invoking foundation models from Amazon and third-party providers through one API, with no infrastructure to manage. SageMaker AI offers flexibility but adds operational overhead, while Comprehend and Polly serve narrow NLP and speech use cases rather than generative model invocation. The unified, serverless model access makes Bedrock the correct choice for a fast prototype.

Answer

Amazon SageMaker AI

Answer

Amazon Polly

Answer

Amazon Comprehend

Question 8

A logistics company is building an internal assistant on Amazon Bedrock that must answer operational questions using its private runbooks and must return citations so staff can verify answers. The team also needs to control cost by limiting how much source text is sent with each request. Which TWO capabilities should they combine to meet these requirements? (Choose two.)

Accepted Answer

Knowledge Bases for Amazon Bedrock to ingest the runbooks and retrieve relevant passages.. Retrieval is the mechanism that both grounds answers in private runbooks and bounds prompt size, because only the most relevant chunks are injected. Knowledge Bases for Amazon Bedrock performs that ingestion and retrieval, and it can return citations linking generated text to source chunks, enabling verification. Capacity reservation, output-length increases, and profanity filtering do not supply grounding, traceability, or cost control.

Answer

Guardrails for Amazon Bedrock with a profanity filter enabled.

Answer

Provisioned Throughput to guarantee dedicated model capacity.

Answer

A larger maxTokens value to include more of each runbook.

Question 9

A retail company wants to build a system that predicts next month's sales for each of its 500 stores based on historical sales, local holidays, and marketing spend. The target values are continuous dollar amounts, and the company has labeled historical data for every store. Which type of machine learning problem does this represent?

Accepted Answer

Supervised regression, because the model learns from labeled data to predict a continuous numeric value.. The company holds labeled historical data and needs a continuous numeric output, next month's sales in dollars. That combination defines supervised regression, where the model learns a mapping from input features to a numeric target. Classification, unsupervised learning, and reinforcement learning all misalign with either the presence of labels or the continuous nature of the predicted value.

Answer

Unsupervised learning, because the model discovers hidden groupings of stores on its own.

Answer

Supervised classification, because the model assigns each store to a predicted category.

Answer

Reinforcement learning, because the model improves by receiving rewards after each prediction.

Question 10

A developer is prompting a foundation model to classify customer feedback into categories. The model sometimes returns extra commentary along with the category label, breaking downstream parsing. The developer wants more deterministic, tightly formatted output without retraining the model. Which technique best addresses this?

Accepted Answer

Set the temperature to zero and constrain the output using a structured format specification.. Setting temperature to zero makes generation greedy and deterministic, while a structured output specification constrains the response to an exact schema. This combination eliminates extraneous commentary and guarantees a parseable label, achieving the desired formatting control without any model retraining. It directly targets both the randomness and the format-enforcement gaps causing the parsing failures.

Answer

Add more few-shot examples that include verbose explanations of each category.

Answer

Increase the maximum output token count to give the model more room to respond.

Answer

Increase the top-p sampling value to broaden the token selection pool.

Question 11

A hospital wants an AI system that reviews chest X-ray images and flags those likely to contain pneumonia so radiologists can prioritize their queue. The hospital has 40,000 historical X-rays, each already labeled by a radiologist as pneumonia present or absent. Which learning approach best fits this scenario?

Accepted Answer

Supervised learning using a classification model trained on the labeled images.. The scenario provides thousands of images already annotated by radiologists with a binary outcome. Supervised classification consumes those labels to learn a decision boundary between pneumonia present and absent, producing the flagging behavior the hospital wants. Unsupervised, reinforcement, and generative approaches either ignore the labels or address a different problem than binary image triage.

Answer

Reinforcement learning where the model is rewarded for each correct flag.

Answer

Generative AI using a foundation model to synthesize new X-ray images for training.

Answer

Unsupervised learning using clustering to group similar X-rays together.

Question 12

A hospital wants to build a model that predicts whether a patient has a specific disease based on labeled historical medical records where each record is marked either positive or negative. Which type of machine learning problem does this represent?

Accepted Answer

Supervised learning using classification. Because each historical record already includes a known positive or negative label and the target is a discrete category, the task is supervised classification. Regression would predict a continuous value, clustering ignores the labels, and reinforcement learning requires rewards from interaction. Classification directly models the disease outcome the hospital wants to predict.

Answer

Reinforcement learning using reward signals

Answer

Supervised learning using regression

Answer

Unsupervised learning using clustering

Question 13

A data scientist trains a model to predict whether a loan applicant will default. After deployment, the model performs well on applicants similar to the training data but poorly on applicants from a newly added geographic region that was underrepresented in training. Which statement best describes the underlying problem?

Accepted Answer

The training data does not represent the new region, so the model cannot generalize to that subpopulation.. The model fails specifically on a subpopulation that was underrepresented in training, which is a data coverage and distribution shift problem. Good performance on familiar applicants confirms the algorithm works, but without representative training examples from the new region the model cannot generalize there. The remedy is additional representative data or techniques that address distribution shift.

Answer

The model has overfit the training set because it memorized noise in the original regions.

Answer

The model requires more training epochs to converge on the new region's data.

Answer

The evaluation metric used is inappropriate and should be replaced with accuracy.

Question 14

A company is building a model to detect fraudulent transactions. The dataset has 1,000,000 transactions, of which only 1,000 are fraudulent. The team wants to evaluate the model's performance. Which metric is most appropriate to use as the primary evaluation metric?

Accepted Answer

F1 score. For highly imbalanced datasets like fraud detection, accuracy is misleading. The F1 score balances precision and recall, making it suitable when both false positives and false negatives matter. It provides a single metric that reflects the model's ability to correctly identify fraud without being overwhelmed by the majority class.

Answer

Recall

Answer

Accuracy

Answer

Precision

Question 15

A financial institution wants to predict whether a loan applicant will default. They have a historical dataset with loan outcomes (default or no default) and various applicant features. Which type of machine learning should they use?

Accepted Answer

Supervised learning. Supervised learning is ideal when historical labeled data is available and the goal is to predict a known outcome. The loan default prediction task is a binary classification problem, which supervised learning handles by learning from past examples to classify new applicants.

Answer

Reinforcement learning

Answer

Unsupervised learning

Answer

Clustering

Question 16

A retail company is preparing to launch a generative AI customer support assistant built on Amazon Bedrock. Before launch, the responsible AI review board asks the team to document the assistant's intended purpose, its known limitations, and the evaluation results from fairness and accuracy testing. Which AWS resource should the team produce to satisfy this request?

Accepted Answer

An Amazon SageMaker Model Card that records intended use, risk rating, and evaluation results for the model.. A responsible AI review board wants structured documentation of purpose, limitations, and evaluation outcomes. SageMaker Model Cards are purpose-built for this, capturing intended use, out-of-scope applications, risk rating, and performance and fairness metrics in a consistent format. The other artifacts cover AWS compliance posture, runtime telemetry, or account advisories, none of which describe the model itself.

Answer

An Amazon CloudWatch dashboard showing invocation counts, latency percentiles, and error rates for the assistant.

Answer

An AWS Trusted Advisor check report listing cost optimization and security recommendations for the account.

Answer

An AWS Artifact report downloaded from the compliance portal covering the AWS services in use.

Question 17

A data science team is preparing a dataset for training a machine learning model. They need to perform data preprocessing to improve model performance. Which TWO of the following are common data preprocessing techniques? (Choose two.)

Accepted Answer

One-hot encoding. Normalization and one-hot encoding are standard preprocessing techniques. Normalization scales numerical features, while one-hot encoding transforms categorical variables into a numerical format. Both are applied to the data before training to ensure the model can effectively learn from all features.

Answer

Feature selection

Answer

Cross-validation

Answer

Gradient descent

Question 18

A marketing team wants to build a generative AI application that produces short promotional copy and accompanying images for product launches. They are evaluating Amazon Bedrock and want to understand which statements accurately describe its capabilities for this use case. (Choose two.)

Accepted Answer

Amazon Bedrock supports both text generation models and image generation models.. Amazon Bedrock offers a unified API across many foundation models and includes both text and image generation capabilities, which fits a promotional copy and imagery workload. It is serverless with respect to model hosting, is available in many Regions, and does not guarantee legal compliance, so the two accurate capability statements are the multi-provider single API and support for text and image models.

Answer

Amazon Bedrock requires customers to provision and manage the underlying GPU infrastructure for each model.

Answer

Amazon Bedrock only supports models hosted in a single AWS Region worldwide.

Answer

Amazon Bedrock automatically guarantees that generated promotional claims are legally compliant.

Question 19

A startup wants to build a mobile app that generates personalized workout plans using a foundation model. They need to minimize infrastructure management and pay only for what they use. Which AWS service should they use to access foundation models via a single API?

Accepted Answer

Amazon Bedrock. The startup needs a managed service that offers foundation models via a single API with pay-as-you-go pricing. Amazon Bedrock provides exactly that, allowing developers to experiment with and deploy models from multiple providers without managing infrastructure. Other services like SageMaker require more management, while Comprehend and Polly serve different purposes.

Answer

Amazon Polly

Answer

Amazon Comprehend

Answer

Amazon SageMaker

Question 20

A media company has millions of customer support emails but no labels indicating topic or sentiment. The company wants to discover natural groupings of emails and reduce dimensionality before further analysis. Which approach should they use?

Accepted Answer

Apply unsupervised learning techniques such as clustering and dimensionality reduction. The absence of labels rules out supervised methods and points to unsupervised learning. Clustering reveals natural groupings among the emails, while dimensionality reduction techniques help compress and visualize the feature space. Together they satisfy both stated goals without requiring any annotation effort.

Answer

Use reinforcement learning to reward correct email topic assignments

Answer

Perform regression on email length to predict customer churn

Answer

Train a supervised classifier on the emails using sentiment labels