AIF-C01 Practice Test 13 — 15 Questions

Question 1

A company wants to generate product descriptions from a few keywords without managing infrastructure. Which AWS service provides a serverless API for accessing foundation models?

Accepted Answer

Amazon Bedrock. Amazon Bedrock is a fully managed, serverless service that provides a single API to access and invoke foundation models (FMs) from leading AI providers like AI21 Labs, Anthropic, Cohere, Meta, and Stability AI. It eliminates the need to manage underlying infrastructure, making it the correct choice for generating product descriptions from keywords without provisioning servers.

Answer

Amazon Lex

Answer

Amazon SageMaker

Answer

Amazon Comprehend

Question 2

A company is building a chatbot that must provide accurate answers based on internal documents without retraining the model. Which approach should they use?

Accepted Answer

Prompt engineering with retrieval-augmented generation (RAG). Retrieval-augmented generation (RAG) allows the chatbot to fetch relevant internal documents at inference time and incorporate them into the prompt, providing accurate, up-to-date answers without retraining the model. This approach combines prompt engineering with a retrieval step, ensuring the model's responses are grounded in the company's specific knowledge base while keeping the base model frozen.

Answer

Reinforcement learning from human feedback (RLHF)

Answer

Fine-tuning the model on internal documents

Answer

Model distillation to a smaller model

Question 3

A developer is building an Amazon Bedrock Agent to automate multi‑step workflows like booking flights and hotels. The Agent needs to call an external flight reservation API. How should the developer expose this API to the Agent?

Accepted Answer

Create an action group that includes an OpenAPI schema describing the flight API and a Lambda function to invoke it. Amazon Bedrock Agents use action groups to define external APIs that the agent can invoke. An action group requires an OpenAPI schema (to describe the API operations and parameters) and a Lambda function (to execute the actual API call). This allows the agent to dynamically select and call the flight reservation API during multi-step workflows.

Answer

Add the API documentation as a context document in the Agent's instructions

Answer

Configure the Agent to use Amazon Kendra to index the API documentation

Answer

Hard‑code the API endpoint in the Agent's prompt using a system message

Question 4

A company uses Amazon Bedrock to generate product descriptions. They want to ensure the outputs consistently follow a specific brand tone (professional yet friendly). They have a small set of example descriptions (few-shot examples) but do not want to fine-tune the model. Which strategy best achieves consistent tone without modifying the base model?

Accepted Answer

Use a system prompt that defines the brand tone and include few-shot examples in the prompt.. Using a system prompt to define the brand tone and including few-shot examples in the prompt is the best strategy because it guides the model's behavior without modifying its weights. This leverages in-context learning, which is effective for style adaptation. Fine-tuning is unnecessary and more resource-intensive.

Answer

Fine-tune the model on a dataset of product descriptions that exemplify the desired tone.

Answer

Implement prompt chaining by breaking the task into multiple steps, each with its own prompt.

Answer

Use retrieval-augmented generation (RAG) to pull example descriptions from a database and prepend them to the prompt.

Question 5

An e-commerce company is building a product description generator using Amazon Bedrock. They want to ensure that the generated descriptions do not include any prohibited content (e.g., offensive language or competitor mentions). The company has a list of denied topics and keywords. Which feature should they use?

Accepted Answer

Bedrock Guardrails with content filters and denied topics. Bedrock Guardrails provides content filters and topic denial that can block specific words, phrases, or entire topics. This is the managed way to enforce content policies on model outputs.

Answer

Amazon Comprehend for toxicity detection

Answer

Bedrock Knowledge Bases with metadata filtering

Answer

Bedrock Agents with a custom action group to filter outputs

Question 6

A team is developing an AI system and wants to document key information such as intended use, performance benchmarks, and limitations. According to AWS best practices for responsible AI, what should they create?

Accepted Answer

Model cards. Model cards are a structured documentation framework recommended by AWS for responsible AI. They provide a standardized way to communicate key information such as intended use, performance benchmarks, limitations, and ethical considerations, ensuring transparency and accountability.

Answer

A whitepaper

Answer

A business requirement document

Answer

A technical blog

Question 7

A data scientist runs the SageMaker Clarify job shown in the exhibit for a credit risk model. After reviewing the results, they find a high bias metric for the gender facet. Which action is most consistent with responsible AI?

Accepted Answer

Investigate the root cause and retrain with balanced data. Responsible AI requires understanding and mitigating bias at its source, not just masking it. Investigating the root cause (e.g., data collection bias, labeling bias, or proxy features) and retraining with balanced data directly addresses the high bias metric detected by SageMaker Clarify, aligning with AWS's principle of fairness. Simply removing the gender attribute may not eliminate bias if other features act as proxies, and increasing the threshold does not fix the underlying model bias.

Answer

Proceed with deployment because the model is already in production

Answer

Remove the gender attribute from the training data and retrain

Answer

Increase the acceptance threshold for the model

Question 8

A company is using Amazon SageMaker to train a large language model with hundreds of billions of parameters. The model does not fit into the memory of a single GPU. Which approach should they use to train the model efficiently?

Accepted Answer

Use SageMaker's model parallelism strategy with the SageMaker distributed training library. SageMaker's model parallelism strategy with the SageMaker distributed training library is specifically designed for training large models that do not fit into the memory of a single GPU. It partitions the model layers across multiple GPUs, enabling efficient training of models with hundreds of billions of parameters by overlapping computation and communication.

Answer

Use a larger instance with more GPU memory, such as p4d.24xlarge

Answer

Use SageMaker's data parallelism strategy

Answer

Reduce the model size by pruning layers until it fits into memory

Question 9

Refer to the exhibit. This is an Amazon Bedrock invocation request for Claude. What is the purpose of the "stop_sequences" parameter?

Accepted Answer

It tells the model to stop generating when it encounters that sequence. The 'stop_sequences' parameter in Amazon Bedrock's invocation request for Claude tells the model to halt generation as soon as it encounters a specified character sequence. This allows developers to control the output format, such as stopping at a newline or a custom delimiter, ensuring the response ends exactly where intended.

Answer

It specifies a character sequence for the model to include in its response

Answer

It limits the number of tokens in the response

Answer

It controls the randomness of the response

Question 10

A data scientist wants to restrict which IAM roles can invoke a specific Amazon Bedrock base model. Which AWS feature should they use?

Accepted Answer

Bedrock resource-based policy. Amazon Bedrock supports resource-based policies that can be attached to a specific base model or custom model to control which IAM principals (roles, users, accounts) can invoke it. This is the intended mechanism for restricting invocation of a particular Bedrock model to a defined set of IAM roles.

Answer

S3 bucket policy on the model artifacts

Answer

AWS KMS key policy for the encryption key

Answer

AWS CloudTrail log delivery policy

Question 11

Which TWO AWS services can be used together to build a chatbot that leverages a foundation model for natural language understanding?

Accepted Answer

Amazon Lex. Amazon Lex (B) is correct because it provides the conversational interface—intents, utterances, slots, and dialog management—needed to build a chatbot that interprets user input. Amazon Bedrock (E) is correct because it offers access to foundation models via a managed API, supplying the natural language understanding and generative responses the chatbot requires; Lex can invoke Bedrock through a Lambda fulfillment function to combine both capabilities. Amazon Rekognition (A) is a computer-vision service for image and video analysis, not natural language understanding, so it does not fit. Amazon Polly (C) only converts text to speech for voice output and does not provide NLU or foundation-model reasoning. AWS Glue (D) is a serverless ETL and data-catalog service for analytics pipelines, unrelated to building conversational chatbots.

Answer

Amazon Rekognition

Answer

Amazon Polly

Answer

AWS Glue

Question 12

Which THREE are key capabilities of Amazon Bedrock? (Choose 3)

Accepted Answer

Model customization through fine-tuning. Amazon Bedrock provides model customization through fine-tuning (B), allowing you to adapt supported foundation models with your own labeled data to improve performance for domain-specific tasks. It also offers Guardrails for Amazon Bedrock (C), which let you define policies that filter harmful or inappropriate content and enforce topics and sensitive-information redaction across model responses. Bedrock additionally delivers serverless inference for foundation models (D), so you can invoke models via a managed API without provisioning or managing any underlying infrastructure. Option A is not a Bedrock capability because Bedrock does not automatically choose a model for you; the developer selects the model. Option E is incorrect because Bedrock itself does not include a built-in vector database; knowledge bases for Amazon Bedrock integrate with separate vector stores such as Amazon OpenSearch Serverless or Amazon Aurora.

Answer

Automatic model selection based on use case

Answer

Built-in vector database for knowledge bases

Question 13

A company deployed a chatbot using Amazon Lex integrated with a Lambda function that invokes Claude on Amazon Bedrock. The Lambda function retrieves relevant documents from an Amazon Kendra index to use as context. Users report that the chatbot's responses are often irrelevant or incorrect despite the Kendra index containing accurate information. The logs show that the Lambda function is correctly passing retrieved documents to the model. What is the most likely cause and solution?

Accepted Answer

The chunking strategy for documents is too coarse or inappropriate; refine chunking and use semantic search in Kendra. When retrieved documents are correctly passed to the model but responses are still irrelevant, the problem is almost always upstream in retrieval quality — specifically how documents were chunked and indexed in Kendra. Coarse or poorly aligned chunks dilute the semantic signal, so the model receives context that does not actually answer the query. Refining chunking strategy and enabling semantic search in Kendra directly addresses the root cause.

Answer

Switch to a larger foundation model like Claude 3 Opus

Answer

The model's temperature is set too high; reduce it to 0.1

Answer

The maximum tokens limit is too low; increase it to 4096

Question 14

A company wants to build a system that automatically categorizes customer support tickets into predefined categories (e.g., billing, technical, account). The team has a large dataset of historical tickets with their category labels. Which type of machine learning problem is this?

Accepted Answer

Multi-class classification. This is a multi-class classification problem because the model must assign each support ticket to one of three or more predefined categories (e.g., billing, technical, account). The dataset provides labeled historical tickets, making it a supervised learning task, and the output is a discrete class label from a set of more than two categories, which distinguishes it from binary classification.

Answer

Regression

Answer

Binary classification

Answer

Clustering

Question 15

A financial services company uses Amazon Bedrock to generate investment report summaries. They have strict compliance requirements that the model must not discuss certain topics like insider trading or unapproved financial advice. Which Bedrock feature should they use to deny these topics?

Accepted Answer

Bedrock Guardrails. Bedrock Guardrails include topic denial policies to block specific subjects. Knowledge Bases, Agents, and prompt engineering alone cannot reliably enforce topic restrictions.

Answer

Prompt engineering with negative instructions

Answer

Bedrock Knowledge Bases

Answer

Bedrock Agents