Amazon AWS AIP-C01 GenAI Solution Architecture Practice Test

 

Topic 01 focuses on GenAI Requirements and Solution Architecture for the AWS Certified Generative AI Developer – Professional certification and the AIP-C01 exam, using Amazon Bedrock and AWS generative AI services and architecture scenarios where relevant. For broader exam preparation, review the AWS Certified Generative AI Developer – Professional AIP-C01 Exam Dumps page. Each question includes a concise explanation of the correct answer and the technical reason the other choices are incorrect.

Question 1

The customer-service platform needs an architecture that can match foundation model capabilities, constraints, and deployment patterns to a specific business use case. Which choice best meets that need?

1. retrieval-augmented generation architecture
2. multimodal foundation model selection
3. requirements-to-model fit analysis
4. Amazon Bedrock proof of concept

Correct Answer: C

 

Correct Answer

Answer C is correct because requirements-to-model fit analysis is designed to match foundation model capabilities, constraints, and deployment patterns to a specific business use case. It compares modality, quality, latency, cost, compliance, context, and integration requirements before selecting an architecture.

Incorrect Answers

Answer A is incorrect because retrieval-augmented generation architecture is primarily used to ground model responses in current enterprise knowledge without retraining the foundation model, which is a different requirement from the one being tested.

Answer B is incorrect because multimodal foundation model selection is primarily used to process a workload that combines text with images, audio, or other supported modalities, which is a different requirement from the one being tested.

Answer D is incorrect because Amazon Bedrock proof of concept is primarily used to validate GenAI feasibility, performance characteristics, and business value before full-scale delivery, which is a different requirement from the one being tested.

 

Question 2

The cloud center of excellence must validate GenAI feasibility, performance characteristics, and business value before full-scale delivery. Which approach is the strongest fit?

1. serverless on-demand invocation pattern
2. AWS Well-Architected Generative AI Lens
3. data-residency-aware regional architecture
4. Amazon Bedrock proof of concept

Correct Answer: D

 

Correct Answer

Answer D is correct because Amazon Bedrock proof of concept is designed to validate GenAI feasibility, performance characteristics, and business value before full-scale delivery. It uses managed foundation models and representative application traffic to test assumptions with limited production risk.

Incorrect Answers

Answer A is incorrect because serverless on-demand invocation pattern is primarily used to support bursty or intermittent inference traffic without continuously running application compute, which is a different requirement from the one being tested.

Answer B is incorrect because AWS Well-Architected Generative AI Lens is primarily used to standardize architecture reviews and identify GenAI-specific design risks, which is a different requirement from the one being tested.

Answer C is incorrect because data-residency-aware regional architecture is primarily used to keep regulated data and model interactions within approved geographic boundaries, which is a different requirement from the one being tested.

 

Question 3

The contact-center transformation team needs an architecture that can standardize architecture reviews and identify GenAI-specific design risks. Which choice best meets that need?

1. managed foundation model service on Amazon Bedrock
2. AWS Well-Architected Generative AI Lens
3. event-driven asynchronous GenAI workflow
4. retrieval-augmented generation architecture

Correct Answer: B

 

Correct Answer

Answer B is correct because AWS Well-Architected Generative AI Lens is designed to standardize architecture reviews and identify GenAI-specific design risks. It extends well-architected guidance with questions and practices for responsible, secure, reliable, performant, and cost-aware GenAI systems.

Incorrect Answers

Answer A is incorrect because managed foundation model service on Amazon Bedrock is primarily used to minimize infrastructure management while integrating supported foundation models through managed APIs, which is a different requirement from the one being tested.

Answer C is incorrect because event-driven asynchronous GenAI workflow is primarily used to decouple long-running GenAI processing from interactive request paths, which is a different requirement from the one being tested.

Answer D is incorrect because retrieval-augmented generation architecture is primarily used to ground model responses in current enterprise knowledge without retraining the foundation model, which is a different requirement from the one being tested.

 

Question 4

The customer analytics platform is prioritizing a requirement to ground model responses in current enterprise knowledge without retraining the foundation model. Which implementation is most appropriate?

1. serverless on-demand invocation pattern
2. multimodal foundation model selection
3. retrieval-augmented generation architecture
4. graceful degradation and fallback design

Correct Answer: C

 

Correct Answer

Answer C is correct because retrieval-augmented generation architecture is designed to ground model responses in current enterprise knowledge without retraining the foundation model. It retrieves relevant external context at inference time and supplies it to the model with the user request.

Incorrect Answers

Answer A is incorrect because serverless on-demand invocation pattern is primarily used to support bursty or intermittent inference traffic without continuously running application compute, which is a different requirement from the one being tested.

Answer B is incorrect because multimodal foundation model selection is primarily used to process a workload that combines text with images, audio, or other supported modalities, which is a different requirement from the one being tested.

Answer D is incorrect because graceful degradation and fallback design is primarily used to maintain useful application behavior when the preferred model or dependency is unavailable, which is a different requirement from the one being tested.

 

Question 5

The enterprise finance automation team needs an architecture that can support bursty or intermittent inference traffic without continuously running application compute. Which choice best meets that need?

1. serverless on-demand invocation pattern
2. data-residency-aware regional architecture
3. event-driven asynchronous GenAI workflow
4. requirements-to-model fit analysis

Correct Answer: A

 

Correct Answer

Answer A is correct because serverless on-demand invocation pattern is designed to support bursty or intermittent inference traffic without continuously running application compute. It uses event-driven or request-driven compute such as Lambda around managed model APIs so capacity is consumed when requests arrive.

Incorrect Answers

Answer B is incorrect because data-residency-aware regional architecture is primarily used to keep regulated data and model interactions within approved geographic boundaries, which is a different requirement from the one being tested.

Answer C is incorrect because event-driven asynchronous GenAI workflow is primarily used to decouple long-running GenAI processing from interactive request paths, which is a different requirement from the one being tested.

Answer D is incorrect because requirements-to-model fit analysis is primarily used to match foundation model capabilities, constraints, and deployment patterns to a specific business use case, which is a different requirement from the one being tested.

 

Question 6

The risk and controls team has a design goal to decouple long-running GenAI processing from interactive request paths. What should the team choose?

1. managed foundation model service on Amazon Bedrock
2. event-driven asynchronous GenAI workflow
3. Amazon Bedrock proof of concept
4. multimodal foundation model selection

Correct Answer: B

 

Correct Answer

Answer B is correct because event-driven asynchronous GenAI workflow is designed to decouple long-running GenAI processing from interactive request paths. It uses queues, events, and workflow orchestration so producers do not wait synchronously for all inference and post-processing steps.

Incorrect Answers

Answer A is incorrect because managed foundation model service on Amazon Bedrock is primarily used to minimize infrastructure management while integrating supported foundation models through managed APIs, which is a different requirement from the one being tested.

Answer C is incorrect because Amazon Bedrock proof of concept is primarily used to validate GenAI feasibility, performance characteristics, and business value before full-scale delivery, which is a different requirement from the one being tested.

Answer D is incorrect because multimodal foundation model selection is primarily used to process a workload that combines text with images, audio, or other supported modalities, which is a different requirement from the one being tested.

 

Question 7

The enterprise search architecture group has a design goal to process a workload that combines text with images, audio, or other supported modalities. What should the team choose?

1. data-residency-aware regional architecture
2. multimodal foundation model selection
3. graceful degradation and fallback design
4. AWS Well-Architected Generative AI Lens

Correct Answer: B

 

Correct Answer

Answer B is correct because multimodal foundation model selection is designed to process a workload that combines text with images, audio, or other supported modalities. It chooses a model whose native modality support matches the input and output types required by the application.

Incorrect Answers

Answer A is incorrect because data-residency-aware regional architecture is primarily used to keep regulated data and model interactions within approved geographic boundaries, which is a different requirement from the one being tested.

Answer C is incorrect because graceful degradation and fallback design is primarily used to maintain useful application behavior when the preferred model or dependency is unavailable, which is a different requirement from the one being tested.

Answer D is incorrect because AWS Well-Architected Generative AI Lens is primarily used to standardize architecture reviews and identify GenAI-specific design risks, which is a different requirement from the one being tested.

 

Question 8

The insurance automation team needs an architecture that can keep regulated data and model interactions within approved geographic boundaries. Which choice best meets that need?

1. data-residency-aware regional architecture
2. managed foundation model service on Amazon Bedrock
3. retrieval-augmented generation architecture
4. requirements-to-model fit analysis

Correct Answer: A

 

Correct Answer

Answer A is correct because data-residency-aware regional architecture is designed to keep regulated data and model interactions within approved geographic boundaries. It selects supported regions, routing, storage, and integration paths that preserve jurisdictional and contractual requirements.

Incorrect Answers

Answer B is incorrect because managed foundation model service on Amazon Bedrock is primarily used to minimize infrastructure management while integrating supported foundation models through managed APIs, which is a different requirement from the one being tested.

Answer C is incorrect because retrieval-augmented generation architecture is primarily used to ground model responses in current enterprise knowledge without retraining the foundation model, which is a different requirement from the one being tested.

Answer D is incorrect because requirements-to-model fit analysis is primarily used to match foundation model capabilities, constraints, and deployment patterns to a specific business use case, which is a different requirement from the one being tested.

 

Question 9

The procurement automation group has a design goal to minimize infrastructure management while integrating supported foundation models through managed APIs. What should the team choose?

1. Amazon Bedrock proof of concept
2. graceful degradation and fallback design
3. serverless on-demand invocation pattern
4. managed foundation model service on Amazon Bedrock

Correct Answer: D

 

Correct Answer

Answer D is correct because managed foundation model service on Amazon Bedrock is designed to minimize infrastructure management while integrating supported foundation models through managed APIs. It provides managed access to multiple foundation models and GenAI features without requiring teams to operate model-serving infrastructure.

Incorrect Answers

Answer A is incorrect because Amazon Bedrock proof of concept is primarily used to validate GenAI feasibility, performance characteristics, and business value before full-scale delivery, which is a different requirement from the one being tested.

Answer B is incorrect because graceful degradation and fallback design is primarily used to maintain useful application behavior when the preferred model or dependency is unavailable, which is a different requirement from the one being tested.

Answer C is incorrect because serverless on-demand invocation pattern is primarily used to support bursty or intermittent inference traffic without continuously running application compute, which is a different requirement from the one being tested.

 

Question 10

The observability engineering group is prioritizing a requirement to maintain useful application behavior when the preferred model or dependency is unavailable. Which implementation is most appropriate?

1. event-driven asynchronous GenAI workflow
2. AWS Well-Architected Generative AI Lens
3. requirements-to-model fit analysis
4. graceful degradation and fallback design

Correct Answer: D

 

Correct Answer

Answer D is correct because graceful degradation and fallback design is designed to maintain useful application behavior when the preferred model or dependency is unavailable. It defines alternate models, reduced-function responses, queues, or deterministic fallbacks instead of failing the whole workflow.

Incorrect Answers

Answer A is incorrect because event-driven asynchronous GenAI workflow is primarily used to decouple long-running GenAI processing from interactive request paths, which is a different requirement from the one being tested.

Answer B is incorrect because AWS Well-Architected Generative AI Lens is primarily used to standardize architecture reviews and identify GenAI-specific design risks, which is a different requirement from the one being tested.

Answer C is incorrect because requirements-to-model fit analysis is primarily used to match foundation model capabilities, constraints, and deployment patterns to a specific business use case, which is a different requirement from the one being tested.

 

Question 11

The claims processing service documents this GenAI behavior: compares modality, quality, latency, cost, compliance, context, and integration requirements before selecting an architecture. Which capability matches it?

1. multimodal foundation model selection
2. retrieval-augmented generation architecture
3. requirements-to-model fit analysis
4. Amazon Bedrock proof of concept

Correct Answer: C

 

Correct Answer

Answer C is correct because the description directly matches requirements-to-model fit analysis. requirements-to-model fit analysis compares modality, quality, latency, cost, compliance, context, and integration requirements before selecting an architecture.

Incorrect Answers

Answer A is incorrect because multimodal foundation model selection is primarily used to process a workload that combines text with images, audio, or other supported modalities, which is a different requirement from the one being tested.

Answer B is incorrect because retrieval-augmented generation architecture is primarily used to ground model responses in current enterprise knowledge without retraining the foundation model, which is a different requirement from the one being tested.

Answer D is incorrect because Amazon Bedrock proof of concept is primarily used to validate GenAI feasibility, performance characteristics, and business value before full-scale delivery, which is a different requirement from the one being tested.

 

Question 12

The clinical knowledge application needs to identify a capability with this behavior: uses managed foundation models and representative application traffic to test assumptions with limited production risk. What is the best match?

1. Amazon Bedrock proof of concept
2. data-residency-aware regional architecture
3. AWS Well-Architected Generative AI Lens
4. serverless on-demand invocation pattern

Correct Answer: A

 

Correct Answer

Answer A is correct because the description directly matches Amazon Bedrock proof of concept. Amazon Bedrock proof of concept uses managed foundation models and representative application traffic to test assumptions with limited production risk.

Incorrect Answers

Answer B is incorrect because data-residency-aware regional architecture is primarily used to keep regulated data and model interactions within approved geographic boundaries, which is a different requirement from the one being tested.

Answer C is incorrect because AWS Well-Architected Generative AI Lens is primarily used to standardize architecture reviews and identify GenAI-specific design risks, which is a different requirement from the one being tested.

Answer D is incorrect because serverless on-demand invocation pattern is primarily used to support bursty or intermittent inference traffic without continuously running application compute, which is a different requirement from the one being tested.

 

Question 13

The mobile experience engineering team documents this GenAI behavior: extends Well-Architected guidance with questions and practices for responsible, secure, reliable, performant, and cost-aware GenAI systems. Which capability matches it?

1. event-driven asynchronous GenAI workflow
2. managed foundation model service on Amazon Bedrock
3. AWS Well-Architected Generative AI Lens
4. retrieval-augmented generation architecture

Correct Answer: C

 

Correct Answer

Answer C is correct because the description directly matches AWS Well-Architected Generative AI Lens. AWS Well-Architected Generative AI Lens extends Well-Architected guidance with questions and practices for responsible, secure, reliable, performant, and cost-aware GenAI systems.

Incorrect Answers

Answer A is incorrect because event-driven asynchronous GenAI workflow is primarily used to decouple long-running GenAI processing from interactive request paths, which is a different requirement from the one being tested.

Answer B is incorrect because managed foundation model service on Amazon Bedrock is primarily used to minimize infrastructure management while integrating supported foundation models through managed APIs, which is a different requirement from the one being tested.

Answer D is incorrect because retrieval-augmented generation architecture is primarily used to ground model responses in current enterprise knowledge without retraining the foundation model, which is a different requirement from the one being tested.

 

Question 14

The cloud application engineering team describes a component that retrieves relevant external context at inference time and supplies it to the model with the user request. Which capability is being described?

1. retrieval-augmented generation architecture
2. multimodal foundation model selection
3. graceful degradation and fallback design
4. serverless on-demand invocation pattern

Correct Answer: A

 

Correct Answer

Answer A is correct because the description directly matches retrieval-augmented generation architecture. retrieval-augmented generation architecture retrieves relevant external context at inference time and supplies it to the model with the user request.

Incorrect Answers

Answer B is incorrect because multimodal foundation model selection is primarily used to process a workload that combines text with images, audio, or other supported modalities, which is a different requirement from the one being tested.

Answer C is incorrect because graceful degradation and fallback design is primarily used to maintain useful application behavior when the preferred model or dependency is unavailable, which is a different requirement from the one being tested.

Answer D is incorrect because serverless on-demand invocation pattern is primarily used to support bursty or intermittent inference traffic without continuously running application compute, which is a different requirement from the one being tested.

 

Question 15

The engineering enablement organization documents this GenAI behavior: uses event-driven or request-driven compute such as Lambda around managed model APIs so capacity is consumed when requests arrive. Which capability matches it?

1. data-residency-aware regional architecture
2. serverless on-demand invocation pattern
3. event-driven asynchronous GenAI workflow
4. requirements-to-model fit analysis

Correct Answer: B

 

Correct Answer

Answer B is correct because the description directly matches serverless on-demand invocation pattern. serverless on-demand invocation pattern uses event-driven or request-driven compute such as Lambda around managed model APIs so capacity is consumed when requests arrive.

Incorrect Answers

Answer A is incorrect because data-residency-aware regional architecture is primarily used to keep regulated data and model interactions within approved geographic boundaries, which is a different requirement from the one being tested.

Answer C is incorrect because event-driven asynchronous GenAI workflow is primarily used to decouple long-running GenAI processing from interactive request paths, which is a different requirement from the one being tested.

Answer D is incorrect because requirements-to-model fit analysis is primarily used to match foundation model capabilities, constraints, and deployment patterns to a specific business use case, which is a different requirement from the one being tested.

 

Question 16

The workflow orchestration team describes a component that uses queues, events, and workflow orchestration so producers do not wait synchronously for all inference and post-processing steps. Which capability is being described?

1. event-driven asynchronous GenAI workflow
2. managed foundation model service on Amazon Bedrock
3. Amazon Bedrock proof of concept
4. multimodal foundation model selection

Correct Answer: A

 

Correct Answer

Answer A is correct because the description directly matches event-driven asynchronous GenAI workflow. event-driven asynchronous GenAI workflow uses queues, events, and workflow orchestration so producers do not wait synchronously for all inference and post-processing steps.

Incorrect Answers

Answer B is incorrect because managed foundation model service on Amazon Bedrock is primarily used to minimize infrastructure management while integrating supported foundation models through managed APIs, which is a different requirement from the one being tested.

Answer C is incorrect because Amazon Bedrock proof of concept is primarily used to validate GenAI feasibility, performance characteristics, and business value before full-scale delivery, which is a different requirement from the one being tested.

Answer D is incorrect because multimodal foundation model selection is primarily used to process a workload that combines text with images, audio, or other supported modalities, which is a different requirement from the one being tested.

 

Question 17

The public-sector application team describes a component that chooses a model whose native modality support matches the input and output types required by the application. Which capability is being described?

1. AWS Well-Architected Generative AI Lens
2. graceful degradation and fallback design
3. multimodal foundation model selection
4. data-residency-aware regional architecture

Correct Answer: C

 

Correct Answer

Answer C is correct because the description directly matches multimodal foundation model selection. multimodal foundation model selection chooses a model whose native modality support matches the input and output types required by the application.

Incorrect Answers

Answer A is incorrect because AWS Well-Architected Generative AI Lens is primarily used to standardize architecture reviews and identify GenAI-specific design risks, which is a different requirement from the one being tested.

Answer B is incorrect because graceful degradation and fallback design is primarily used to maintain useful application behavior when the preferred model or dependency is unavailable, which is a different requirement from the one being tested.

Answer D is incorrect because data-residency-aware regional architecture is primarily used to keep regulated data and model interactions within approved geographic boundaries, which is a different requirement from the one being tested.

 

Question 18

The internal AI platform team describes a component that selects supported Regions, routing, storage, and integration paths that preserve jurisdictional and contractual requirements. Which capability is being described?

1. data-residency-aware regional architecture
2. requirements-to-model fit analysis
3. managed foundation model service on Amazon Bedrock
4. retrieval-augmented generation architecture

Correct Answer: A

 

Correct Answer

Answer A is correct because the description directly matches data-residency-aware regional architecture. data-residency-aware regional architecture selects supported Regions, routing, storage, and integration paths that preserve jurisdictional and contractual requirements.

Incorrect Answers

Answer B is incorrect because requirements-to-model fit analysis is primarily used to match foundation model capabilities, constraints, and deployment patterns to a specific business use case, which is a different requirement from the one being tested.

Answer C is incorrect because managed foundation model service on Amazon Bedrock is primarily used to minimize infrastructure management while integrating supported foundation models through managed APIs, which is a different requirement from the one being tested.

Answer D is incorrect because retrieval-augmented generation architecture is primarily used to ground model responses in current enterprise knowledge without retraining the foundation model, which is a different requirement from the one being tested.

 

Question 19

Within the data privacy office’s architecture, which capability matches this technical description: provides managed access to multiple foundation models and GenAI features without requiring teams to operate model-serving infrastructure?

1. Amazon Bedrock proof of concept
2. graceful degradation and fallback design
3. managed foundation model service on Amazon Bedrock
4. serverless on-demand invocation pattern

Correct Answer: C

 

Correct Answer

Answer C is correct because the description directly matches managed foundation model service on Amazon Bedrock. managed foundation model service on Amazon Bedrock provides managed access to multiple foundation models and GenAI features without requiring teams to operate model-serving infrastructure.

Incorrect Answers

Answer A is incorrect because Amazon Bedrock proof of concept is primarily used to validate GenAI feasibility, performance characteristics, and business value before full-scale delivery, which is a different requirement from the one being tested.

Answer B is incorrect because graceful degradation and fallback design is primarily used to maintain useful application behavior when the preferred model or dependency is unavailable, which is a different requirement from the one being tested.

Answer D is incorrect because serverless on-demand invocation pattern is primarily used to support bursty or intermittent inference traffic without continuously running application compute, which is a different requirement from the one being tested.

 

Question 20

The enterprise integration office describes a component that defines alternate models, reduced-function responses, queues, or deterministic fallbacks instead of failing the whole workflow. Which capability is being described?

1. graceful degradation and fallback design
2. AWS Well-Architected Generative AI Lens
3. requirements-to-model fit analysis
4. event-driven asynchronous GenAI workflow

Correct Answer: A

 

Correct Answer

Answer A is correct because the description directly matches graceful degradation and fallback design. graceful degradation and fallback design defines alternate models, reduced-function responses, queues, or deterministic fallbacks instead of failing the whole workflow.

Incorrect Answers

Answer B is incorrect because AWS Well-Architected Generative AI Lens is primarily used to standardize architecture reviews and identify GenAI-specific design risks, which is a different requirement from the one being tested.

Answer C is incorrect because requirements-to-model fit analysis is primarily used to match foundation model capabilities, constraints, and deployment patterns to a specific business use case, which is a different requirement from the one being tested.

Answer D is incorrect because event-driven asynchronous GenAI workflow is primarily used to decouple long-running GenAI processing from interactive request paths, which is a different requirement from the one being tested.

 

Question 21

Why would the API platform team introduce requirements-to-model fit analysis into the GenAI architecture?

1. Ground model responses in current enterprise knowledge without retraining the foundation model
2. Match foundation model capabilities, constraints, and deployment patterns to a specific business use case
3. Process a workload that combines text with images, audio, or other supported modalities
4. Validate GenAI feasibility, performance characteristics, and business value before full-scale delivery

Correct Answer: B

 

Correct Answer

Answer B is correct because requirements-to-model fit analysis is specifically used to match foundation model capabilities, constraints, and deployment patterns to a specific business use case. It compares modality, quality, latency, cost, compliance, context, and integration requirements before selecting an architecture.

Incorrect Answers

Answer A is incorrect because that requirement aligns with retrieval-augmented generation architecture, not requirements-to-model fit analysis.

Answer C is incorrect because that requirement aligns with multimodal foundation model selection, not requirements-to-model fit analysis.

Answer D is incorrect because that requirement aligns with Amazon Bedrock proof of concept, not requirements-to-model fit analysis.

 

Question 22

The product recommendation platform includes Amazon Bedrock proof of concept in its architecture. What is its primary role?

1. Support bursty or intermittent inference traffic without continuously running application compute
2. Standardize architecture reviews and identify GenAI-specific design risks
3. Keep regulated data and model interactions within approved geographic boundaries
4. Validate GenAI feasibility, performance characteristics, and business value before full-scale delivery

Correct Answer: D

 

Correct Answer

Answer D is correct because Amazon Bedrock proof of concept is specifically used to validate GenAI feasibility, performance characteristics, and business value before full-scale delivery. It uses managed foundation models and representative application traffic to test assumptions with limited production risk.

Incorrect Answers

Answer A is incorrect because that requirement aligns with serverless on-demand invocation pattern, not Amazon Bedrock proof of concept.

Answer B is incorrect because that requirement aligns with AWS Well-Architected Generative AI Lens, not Amazon Bedrock proof of concept.

Answer C is incorrect because that requirement aligns with data-residency-aware regional architecture, not Amazon Bedrock proof of concept.

 

Question 23

The audit analytics group includes AWS Well-Architected Generative AI Lens in its architecture. What is its primary role?

1. Ground model responses in current enterprise knowledge without retraining the foundation model
2. Minimize infrastructure management while integrating supported foundation models through managed APIs
3. Decouple long-running GenAI processing from interactive request paths
4. Standardize architecture reviews and identify GenAI-specific design risks

Correct Answer: D

 

Correct Answer

Answer D is correct because AWS Well-Architected Generative AI Lens is specifically used to standardize architecture reviews and identify GenAI-specific design risks. It extends well-architected guidance with questions and practices for responsible, secure, reliable, performant, and cost-aware GenAI systems.

Incorrect Answers

Answer A is incorrect because that requirement aligns with retrieval-augmented generation architecture, not AWS Well-Architected Generative AI Lens.

Answer B is incorrect because that requirement aligns with managed foundation model service on Amazon Bedrock, not AWS Well-Architected Generative AI Lens.

Answer C is incorrect because that requirement aligns with event-driven asynchronous GenAI workflow, not AWS Well-Architected Generative AI Lens.

 

Question 24

The AI platform governance team is considering retrieval-augmented generation architecture. What problem is this choice primarily meant to solve?

1. Maintain useful application behavior when the preferred model or dependency is unavailable
2. Support bursty or intermittent inference traffic without continuously running application compute
3. Process a workload that combines text with images, audio, or other supported modalities
4. Ground model responses in current enterprise knowledge without retraining the foundation model

Correct Answer: D

 

Correct Answer

Answer D is correct because retrieval-augmented generation architecture is specifically used to ground model responses in current enterprise knowledge without retraining the foundation model. It retrieves relevant external context at inference time and supplies it to the model with the user request.

Incorrect Answers

Answer A is incorrect because that requirement aligns with graceful degradation and fallback design, not retrieval-augmented generation architecture.

Answer B is incorrect because that requirement aligns with serverless on-demand invocation pattern, not retrieval-augmented generation architecture.

Answer C is incorrect because that requirement aligns with multimodal foundation model selection, not retrieval-augmented generation architecture.

 

Question 25

The transaction monitoring platform is considering serverless on-demand invocation pattern. What problem is this choice primarily meant to solve?

1. Match foundation model capabilities, constraints, and deployment patterns to a specific business use case
2. Support bursty or intermittent inference traffic without continuously running application compute
3. Keep regulated data and model interactions within approved geographic boundaries
4. Decouple long-running GenAI processing from interactive request paths

Correct Answer: B

 

Correct Answer

Answer B is correct because serverless on-demand invocation pattern is specifically used to support bursty or intermittent inference traffic without continuously running application compute. It uses event-driven or request-driven compute such as Lambda around managed model APIs so capacity is consumed when requests arrive.

Incorrect Answers

Answer A is incorrect because that requirement aligns with requirements-to-model fit analysis, not serverless on-demand invocation pattern.

Answer C is incorrect because that requirement aligns with data-residency-aware regional architecture, not serverless on-demand invocation pattern.

Answer D is incorrect because that requirement aligns with event-driven asynchronous GenAI workflow, not serverless on-demand invocation pattern.