Amazon AWS AIP-C01 Foundation Model API Integration Practice Test

 

Topic 10 focuses on Foundation Model API Integration and Routing 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 enterprise data marketplace has a design goal to return a model response within the same request-response interaction. What should the team choose?

1. synchronous Amazon Bedrock invocation
2. model fallback mechanism
3. WebSocket delivery
4. AWS SDK with Amazon SQS for asynchronous inference work

Correct Answer: A

 

Correct Answer

Answer A is correct because synchronous Amazon Bedrock invocation is designed to return a model response within the same request-response interaction. It uses a Bedrock runtime API when the caller should wait for the inference result immediately.

Incorrect Answers

Answer B is incorrect because model fallback mechanism is primarily used to continue serving requests when the preferred foundation model is unavailable or fails, which is a different requirement from the one being tested.

Answer C is incorrect because WebSocket delivery is primarily used to maintain a bidirectional real-time connection for interactive AI experiences, which is a different requirement from the one being tested.

Answer D is incorrect because AWS SDK with Amazon SQS for asynchronous inference work is primarily used to decouple callers from long-running or bursty model processing, which is a different requirement from the one being tested.

 

Question 2

The case management platform has a design goal to decouple callers from long-running or bursty model processing. What should the team choose?

1. API Gateway request validation
2. AWS SDK with Amazon SQS for asynchronous inference work
3. AWS X-Ray distributed tracing
4. server-sent events

Correct Answer: B

 

Correct Answer

Answer B is correct because AWS SDK with Amazon SQS for asynchronous inference work is designed to decouple callers from long-running or bursty model processing. It places work on a durable queue and lets workers invoke model APIs independently of the original request.

Incorrect Answers

Answer A is incorrect because API Gateway request validation is primarily used to reject malformed API requests before they reach the GenAI backend, which is a different requirement from the one being tested.

Answer C is incorrect because AWS X-Ray distributed tracing is primarily used to trace model API calls across application and service boundaries, which is a different requirement from the one being tested.

Answer D is incorrect because server-sent events is primarily used to stream one-way incremental model output from server to browser over an HTTP connection, which is a different requirement from the one being tested.

 

Question 3

The legal research platform is prioritizing a requirement to reject malformed API requests before they reach the GenAI backend. Which implementation is most appropriate?

1. Amazon Bedrock streaming response
2. static model routing
3. AWS SDK exponential backoff
4. API Gateway request validation

Correct Answer: D

 

Correct Answer

Answer D is correct because API Gateway request validation is designed to reject malformed API requests before they reach the GenAI backend. It checks required parameters, payload structure, and API-level constraints at the managed gateway.

Incorrect Answers

Answer A is incorrect because Amazon Bedrock streaming response is primarily used to deliver generated output incrementally as tokens or chunks become available, which is a different requirement from the one being tested.

Answer B is incorrect because static model routing is primarily used to send known request categories to predetermined models using simple deterministic logic, which is a different requirement from the one being tested.

Answer C is incorrect because AWS SDK exponential backoff is primarily used to retry transient throttling or service failures without creating a synchronized retry storm, which is a different requirement from the one being tested.

 

Question 4

The research assistant application must deliver generated output incrementally as tokens or chunks become available. Which approach is the strongest fit?

1. Amazon Bedrock streaming response
2. Step Functions dynamic model routing
3. WebSocket delivery
4. API Gateway throttling

Correct Answer: A

 

Correct Answer

Answer A is correct because Amazon Bedrock streaming response is designed to deliver generated output incrementally as tokens or chunks become available. It uses a streaming inference interface so users see partial output before the complete response is finished.

Incorrect Answers

Answer B is incorrect because Step Functions dynamic model routing is primarily used to choose models through multi-step runtime logic based on request content or measured conditions, which is a different requirement from the one being tested.

Answer C is incorrect because WebSocket delivery is primarily used to maintain a bidirectional real-time connection for interactive AI experiences, which is a different requirement from the one being tested.

Answer D is incorrect because API Gateway throttling is primarily used to enforce API-level request-rate limits before model capacity is overwhelmed, which is a different requirement from the one being tested.

 

Question 5

The AI product engineering team must maintain a bidirectional real-time connection for interactive AI experiences. Which approach is the strongest fit?

1. WebSocket delivery
2. synchronous Amazon Bedrock invocation
3. server-sent events
4. model fallback mechanism

Correct Answer: A

 

Correct Answer

Answer A is correct because WebSocket delivery is designed to maintain a bidirectional real-time connection for interactive AI experiences. It keeps a persistent connection through which clients and servers can exchange messages asynchronously.

Incorrect Answers

Answer B is incorrect because synchronous Amazon Bedrock invocation is primarily used to return a model response within the same request-response interaction, which is a different requirement from the one being tested.

Answer C is incorrect because server-sent events is primarily used to stream one-way incremental model output from server to browser over an HTTP connection, which is a different requirement from the one being tested.

Answer D is incorrect because model fallback mechanism is primarily used to continue serving requests when the preferred foundation model is unavailable or fails, which is a different requirement from the one being tested.

 

Question 6

The supply-chain analytics group has a design goal to stream one-way incremental model output from server to browser over an HTTP connection. What should the team choose?

1. AWS SDK exponential backoff
2. AWS SDK with Amazon SQS for asynchronous inference work
3. AWS X-Ray distributed tracing
4. server-sent events

Correct Answer: D

 

Correct Answer

Answer D is correct because server-sent events is designed to stream one-way incremental model output from server to browser over an HTTP connection. It uses an event stream that is simpler than full bidirectional websockets when the client mainly receives updates.

Incorrect Answers

Answer A is incorrect because AWS SDK exponential backoff is primarily used to retry transient throttling or service failures without creating a synchronized retry storm, which is a different requirement from the one being tested.

Answer B is incorrect because AWS SDK with Amazon SQS for asynchronous inference work is primarily used to decouple callers from long-running or bursty model processing, which is a different requirement from the one being tested.

Answer C is incorrect because AWS X-Ray distributed tracing is primarily used to trace model API calls across application and service boundaries, which is a different requirement from the one being tested.

 

Question 7

The enterprise developer platform must retry transient throttling or service failures without creating a synchronized retry storm. Which approach is the strongest fit?

1. static model routing
2. AWS SDK exponential backoff
3. API Gateway request validation
4. API Gateway throttling

Correct Answer: B

 

Correct Answer

Answer B is correct because AWS SDK exponential backoff is designed to retry transient throttling or service failures without creating a synchronized retry storm. It increases delay between repeated attempts and typically adds jitter according to sdk retry behavior.

Incorrect Answers

Answer A is incorrect because static model routing is primarily used to send known request categories to predetermined models using simple deterministic logic, which is a different requirement from the one being tested.

Answer C is incorrect because API Gateway request validation is primarily used to reject malformed API requests before they reach the GenAI backend, which is a different requirement from the one being tested.

Answer D is incorrect because API Gateway throttling is primarily used to enforce API-level request-rate limits before model capacity is overwhelmed, which is a different requirement from the one being tested.

 

Question 8

The customer identity platform must enforce API-level request-rate limits before model capacity is overwhelmed. Which approach is the strongest fit?

1. Amazon Bedrock streaming response
2. model fallback mechanism
3. Step Functions dynamic model routing
4. API Gateway throttling

Correct Answer: D

 

Correct Answer

Answer D is correct because API Gateway throttling is designed to enforce API-level request-rate limits before model capacity is overwhelmed. It applies rate and burst controls to client traffic at the managed API boundary.

Incorrect Answers

Answer A is incorrect because Amazon Bedrock streaming response is primarily used to deliver generated output incrementally as tokens or chunks become available, which is a different requirement from the one being tested.

Answer B is incorrect because model fallback mechanism is primarily used to continue serving requests when the preferred foundation model is unavailable or fails, which is a different requirement from the one being tested.

Answer C is incorrect because Step Functions dynamic model routing is primarily used to choose models through multi-step runtime logic based on request content or measured conditions, which is a different requirement from the one being tested.

 

Question 9

The business intelligence application has a design goal to continue serving requests when the preferred foundation model is unavailable or fails. What should the team choose?

1. AWS X-Ray distributed tracing
2. WebSocket delivery
3. model fallback mechanism
4. synchronous Amazon Bedrock invocation

Correct Answer: C

 

Correct Answer

Answer C is correct because model fallback mechanism is designed to continue serving requests when the preferred foundation model is unavailable or fails. It routes to an alternate model or reduced-function response according to predefined failure rules.

Incorrect Answers

Answer A is incorrect because AWS X-Ray distributed tracing is primarily used to trace model API calls across application and service boundaries, which is a different requirement from the one being tested.

Answer B is incorrect because WebSocket delivery is primarily used to maintain a bidirectional real-time connection for interactive AI experiences, which is a different requirement from the one being tested.

Answer D is incorrect because synchronous Amazon Bedrock invocation is primarily used to return a model response within the same request-response interaction, which is a different requirement from the one being tested.

 

Question 10

The healthcare document platform must trace model API calls across application and service boundaries. Which approach is the strongest fit?

1. AWS SDK with Amazon SQS for asynchronous inference work
2. server-sent events
3. AWS X-Ray distributed tracing
4. static model routing

Correct Answer: C

 

Correct Answer

Answer C is correct because AWS X-Ray distributed tracing is designed to trace model API calls across application and service boundaries. It records trace segments so latency and failures can be correlated across a distributed request path.

Incorrect Answers

Answer A is incorrect because AWS SDK with Amazon SQS for asynchronous inference work is primarily used to decouple callers from long-running or bursty model processing, which is a different requirement from the one being tested.

Answer B is incorrect because server-sent events is primarily used to stream one-way incremental model output from server to browser over an HTTP connection, which is a different requirement from the one being tested.

Answer D is incorrect because static model routing is primarily used to send known request categories to predetermined models using simple deterministic logic, which is a different requirement from the one being tested.

 

Question 11

The e-commerce search team is prioritizing a requirement to send known request categories to predetermined models using simple deterministic logic. Which implementation is most appropriate?

1. Step Functions dynamic model routing
2. AWS SDK exponential backoff
3. API Gateway request validation
4. static model routing

Correct Answer: D

 

Correct Answer

Answer D is correct because static model routing is designed to send known request categories to predetermined models using simple deterministic logic. It uses configuration or application rules when model choice does not need runtime inference or dynamic orchestration.

Incorrect Answers

Answer A is incorrect because Step Functions dynamic model routing is primarily used to choose models through multi-step runtime logic based on request content or measured conditions, which is a different requirement from the one being tested.

Answer B is incorrect because AWS SDK exponential backoff is primarily used to retry transient throttling or service failures without creating a synchronized retry storm, which is a different requirement from the one being tested.

Answer C is incorrect because API Gateway request validation is primarily used to reject malformed API requests before they reach the GenAI backend, which is a different requirement from the one being tested.

 

Question 12

The risk analytics team is prioritizing a requirement to choose models through multi-step runtime logic based on request content or measured conditions. Which implementation is most appropriate?

1. API Gateway throttling
2. Step Functions dynamic model routing
3. synchronous Amazon Bedrock invocation
4. Amazon Bedrock streaming response

Correct Answer: B

 

Correct Answer

Answer B is correct because Step Functions dynamic model routing is designed to choose models through multi-step runtime logic based on request content or measured conditions. It encodes branching and orchestration in a durable workflow instead of hard-coding every routing path in one service.

Incorrect Answers

Answer A is incorrect because API Gateway throttling is primarily used to enforce API-level request-rate limits before model capacity is overwhelmed, which is a different requirement from the one being tested.

Answer C is incorrect because synchronous Amazon Bedrock invocation is primarily used to return a model response within the same request-response interaction, which is a different requirement from the one being tested.

Answer D is incorrect because Amazon Bedrock streaming response is primarily used to deliver generated output incrementally as tokens or chunks become available, which is a different requirement from the one being tested.

 

Question 13

The education technology platform needs to identify a capability with this behavior: uses a Bedrock runtime API when the caller should wait for the inference result immediately. What is the best match?

1. WebSocket delivery
2. model fallback mechanism
3. synchronous Amazon Bedrock invocation
4. AWS SDK with Amazon SQS for asynchronous inference work

Correct Answer: C

 

Correct Answer

Answer C is correct because the description directly matches synchronous Amazon Bedrock invocation. synchronous Amazon Bedrock invocation uses a Bedrock runtime API when the caller should wait for the inference result immediately.

Incorrect Answers

Answer A is incorrect because WebSocket delivery is primarily used to maintain a bidirectional real-time connection for interactive AI experiences, which is a different requirement from the one being tested.

Answer B is incorrect because model fallback mechanism is primarily used to continue serving requests when the preferred foundation model is unavailable or fails, which is a different requirement from the one being tested.

Answer D is incorrect because AWS SDK with Amazon SQS for asynchronous inference work is primarily used to decouple callers from long-running or bursty model processing, which is a different requirement from the one being tested.

 

Question 14

The field-service application describes a component that places work on a durable queue and lets workers invoke model APIs independently of the original request. Which capability is being described?

1. AWS SDK with Amazon SQS for asynchronous inference work
2. AWS X-Ray distributed tracing
3. server-sent events
4. API Gateway request validation

Correct Answer: A

 

Correct Answer

Answer A is correct because the description directly matches AWS SDK with Amazon SQS for asynchronous inference work. AWS SDK with Amazon SQS for asynchronous inference work places work on a durable queue and lets workers invoke model APIs independently of the original request.

Incorrect Answers

Answer B is incorrect because AWS X-Ray distributed tracing is primarily used to trace model API calls across application and service boundaries, which is a different requirement from the one being tested.

Answer C is incorrect because server-sent events is primarily used to stream one-way incremental model output from server to browser over an HTTP connection, which is a different requirement from the one being tested.

Answer D is incorrect because API Gateway request validation is primarily used to reject malformed API requests before they reach the GenAI backend, which is a different requirement from the one being tested.

 

Question 15

The model governance committee documents this GenAI behavior: checks required parameters, payload structure, and API-level constraints at the managed gateway. Which capability matches it?

1. Amazon Bedrock streaming response
2. API Gateway request validation
3. static model routing
4. AWS SDK exponential backoff

Correct Answer: B

 

Correct Answer

Answer B is correct because the description directly matches API Gateway request validation. API Gateway request validation checks required parameters, payload structure, and API-level constraints at the managed gateway.

Incorrect Answers

Answer A is incorrect because Amazon Bedrock streaming response is primarily used to deliver generated output incrementally as tokens or chunks become available, which is a different requirement from the one being tested.

Answer C is incorrect because static model routing is primarily used to send known request categories to predetermined models using simple deterministic logic, which is a different requirement from the one being tested.

Answer D is incorrect because AWS SDK exponential backoff is primarily used to retry transient throttling or service failures without creating a synchronized retry storm, which is a different requirement from the one being tested.

 

Question 16

Within the AI governance program’s architecture, which capability matches this technical description: uses a streaming inference interface so users see partial output before the complete response is finished?

1. API Gateway throttling
2. Step Functions dynamic model routing
3. WebSocket delivery
4. Amazon Bedrock streaming response

Correct Answer: D

 

Correct Answer

Answer D is correct because the description directly matches Amazon Bedrock streaming response. Amazon Bedrock streaming response uses a streaming inference interface so users see partial output before the complete response is finished.

Incorrect Answers

Answer A is incorrect because API Gateway throttling is primarily used to enforce API-level request-rate limits before model capacity is overwhelmed, which is a different requirement from the one being tested.

Answer B is incorrect because Step Functions dynamic model routing is primarily used to choose models through multi-step runtime logic based on request content or measured conditions, which is a different requirement from the one being tested.

Answer C is incorrect because WebSocket delivery is primarily used to maintain a bidirectional real-time connection for interactive AI experiences, which is a different requirement from the one being tested.

 

Question 17

Within the sales operations automation team’s architecture, which capability matches this technical description: keeps a persistent connection through which clients and servers can exchange messages asynchronously?

1. WebSocket delivery
2. synchronous Amazon Bedrock invocation
3. server-sent events
4. model fallback mechanism

Correct Answer: A

 

Correct Answer

Answer A is correct because the description directly matches WebSocket delivery. WebSocket delivery keeps a persistent connection through which clients and servers can exchange messages asynchronously.

Incorrect Answers

Answer B is incorrect because synchronous Amazon Bedrock invocation is primarily used to return a model response within the same request-response interaction, which is a different requirement from the one being tested.

Answer C is incorrect because server-sent events is primarily used to stream one-way incremental model output from server to browser over an HTTP connection, which is a different requirement from the one being tested.

Answer D is incorrect because model fallback mechanism is primarily used to continue serving requests when the preferred foundation model is unavailable or fails, which is a different requirement from the one being tested.

 

Question 18

Within the knowledge retrieval engineering team’s architecture, which capability matches this technical description: uses an event stream that is simpler than full bidirectional WebSockets when the client mainly receives updates?

1. AWS SDK exponential backoff
2. AWS X-Ray distributed tracing
3. server-sent events
4. AWS SDK with Amazon SQS for asynchronous inference work

Correct Answer: C

 

Correct Answer

Answer C is correct because the description directly matches server-sent events. server-sent events uses an event stream that is simpler than full bidirectional WebSockets when the client mainly receives updates.

Incorrect Answers

Answer A is incorrect because AWS SDK exponential backoff is primarily used to retry transient throttling or service failures without creating a synchronized retry storm, which is a different requirement from the one being tested.

Answer B is incorrect because AWS X-Ray distributed tracing is primarily used to trace model API calls across application and service boundaries, which is a different requirement from the one being tested.

Answer D is incorrect because AWS SDK with Amazon SQS for asynchronous inference work is primarily used to decouple callers from long-running or bursty model processing, which is a different requirement from the one being tested.

 

Question 19

Within the enterprise content repository team’s architecture, which capability matches this technical description: increases delay between repeated attempts and typically adds jitter according to SDK retry behavior?

1. AWS SDK exponential backoff
2. API Gateway request validation
3. API Gateway throttling
4. static model routing

Correct Answer: A

 

Correct Answer

Answer A is correct because the description directly matches AWS SDK exponential backoff. AWS SDK exponential backoff increases delay between repeated attempts and typically adds jitter according to SDK retry behavior.

Incorrect Answers

Answer B is incorrect because API Gateway request validation is primarily used to reject malformed API requests before they reach the GenAI backend, which is a different requirement from the one being tested.

Answer C is incorrect because API Gateway throttling is primarily used to enforce API-level request-rate limits before model capacity is overwhelmed, which is a different requirement from the one being tested.

Answer D is incorrect because static model routing is primarily used to send known request categories to predetermined models using simple deterministic logic, which is a different requirement from the one being tested.

 

Question 20

The manufacturing analytics team needs to identify a capability with this behavior: applies rate and burst controls to client traffic at the managed API boundary. What is the best match?

1. model fallback mechanism
2. Step Functions dynamic model routing
3. Amazon Bedrock streaming response
4. API Gateway throttling

Correct Answer: D

 

Correct Answer

Answer D is correct because the description directly matches API Gateway throttling. API Gateway throttling applies rate and burst controls to client traffic at the managed API boundary.

Incorrect Answers

Answer A is incorrect because model fallback mechanism is primarily used to continue serving requests when the preferred foundation model is unavailable or fails, which is a different requirement from the one being tested.

Answer B is incorrect because Step Functions dynamic model routing is primarily used to choose models through multi-step runtime logic based on request content or measured conditions, which is a different requirement from the one being tested.

Answer C is incorrect because Amazon Bedrock streaming response is primarily used to deliver generated output incrementally as tokens or chunks become available, which is a different requirement from the one being tested.

 

Question 21

The developer productivity team describes a component that routes to an alternate model or reduced-function response according to predefined failure rules. Which capability is being described?

1. AWS X-Ray distributed tracing
2. synchronous Amazon Bedrock invocation
3. model fallback mechanism
4. WebSocket delivery

Correct Answer: C

 

Correct Answer

Answer C is correct because the description directly matches model fallback mechanism. model fallback mechanism routes to an alternate model or reduced-function response according to predefined failure rules.

Incorrect Answers

Answer A is incorrect because AWS X-Ray distributed tracing is primarily used to trace model API calls across application and service boundaries, which is a different requirement from the one being tested.

Answer B is incorrect because synchronous Amazon Bedrock invocation is primarily used to return a model response within the same request-response interaction, which is a different requirement from the one being tested.

Answer D is incorrect because WebSocket delivery is primarily used to maintain a bidirectional real-time connection for interactive AI experiences, which is a different requirement from the one being tested.

 

Question 22

The business process automation team needs to identify a capability with this behavior: records trace segments so latency and failures can be correlated across a distributed request path. What is the best match?

1. AWS SDK with Amazon SQS for asynchronous inference work
2. AWS X-Ray distributed tracing
3. server-sent events
4. static model routing

Correct Answer: B

 

Correct Answer

Answer B is correct because the description directly matches AWS X-Ray distributed tracing. AWS X-Ray distributed tracing records trace segments so latency and failures can be correlated across a distributed request path.

Incorrect Answers

Answer A is incorrect because AWS SDK with Amazon SQS for asynchronous inference work is primarily used to decouple callers from long-running or bursty model processing, which is a different requirement from the one being tested.

Answer C is incorrect because server-sent events is primarily used to stream one-way incremental model output from server to browser over an HTTP connection, which is a different requirement from the one being tested.

Answer D is incorrect because static model routing is primarily used to send known request categories to predetermined models using simple deterministic logic, which is a different requirement from the one being tested.

 

Question 23

The regulatory reporting platform needs to identify a capability with this behavior: uses configuration or application rules when model choice does not need runtime inference or dynamic orchestration. What is the best match?

1. static model routing
2. API Gateway request validation
3. AWS SDK exponential backoff
4. Step Functions dynamic model routing

Correct Answer: A

 

Correct Answer

Answer A is correct because the description directly matches static model routing. static model routing uses configuration or application rules when model choice does not need runtime inference or dynamic orchestration.

Incorrect Answers

Answer B is incorrect because API Gateway request validation is primarily used to reject malformed API requests before they reach the GenAI backend, which is a different requirement from the one being tested.

Answer C is incorrect because AWS SDK exponential backoff is primarily used to retry transient throttling or service failures without creating a synchronized retry storm, which is a different requirement from the one being tested.

Answer D is incorrect because Step Functions dynamic model routing is primarily used to choose models through multi-step runtime logic based on request content or measured conditions, which is a different requirement from the one being tested.

 

Question 24

The privacy engineering function needs to identify a capability with this behavior: encodes branching and orchestration in a durable workflow instead of hard-coding every routing path in one service. What is the best match?

1. API Gateway throttling
2. Step Functions dynamic model routing
3. Amazon Bedrock streaming response
4. synchronous Amazon Bedrock invocation

Correct Answer: B

 

Correct Answer

Answer B is correct because the description directly matches Step Functions dynamic model routing. Step Functions dynamic model routing encodes branching and orchestration in a durable workflow instead of hard-coding every routing path in one service.

Incorrect Answers

Answer A is incorrect because API Gateway throttling is primarily used to enforce API-level request-rate limits before model capacity is overwhelmed, which is a different requirement from the one being tested.

Answer C is incorrect because Amazon Bedrock streaming response is primarily used to deliver generated output incrementally as tokens or chunks become available, which is a different requirement from the one being tested.

Answer D is incorrect because synchronous Amazon Bedrock invocation is primarily used to return a model response within the same request-response interaction, which is a different requirement from the one being tested.

 

Question 25

The contact-center transformation team has proposed synchronous Amazon Bedrock invocation for its design. Which requirement best justifies it?

1. Decouple callers from long-running or bursty model processing
2. Return a model response within the same request-response interaction
3. Maintain a bidirectional real-time connection for interactive AI experiences
4. Continue serving requests when the preferred foundation model is unavailable or fails

Correct Answer: B

 

Correct Answer

Answer B is correct because synchronous Amazon Bedrock invocation is specifically used to return a model response within the same request-response interaction. It uses a Bedrock runtime API when the caller should wait for the inference result immediately.

Incorrect Answers

Answer A is incorrect because that requirement aligns with AWS SDK with Amazon SQS for asynchronous inference work, not synchronous Amazon Bedrock invocation.

Answer C is incorrect because that requirement aligns with WebSocket delivery, not synchronous Amazon Bedrock invocation.

Answer D is incorrect because that requirement aligns with model fallback mechanism, not synchronous Amazon Bedrock invocation.

 

Question 26

The customer analytics platform includes AWS SDK with Amazon SQS for asynchronous inference work in its architecture. What is its primary role?

1. Trace model API calls across application and service boundaries
2. Decouple callers from long-running or bursty model processing
3. Reject malformed API requests before they reach the GenAI backend
4. Stream one-way incremental model output from server to browser over an HTTP connection

Correct Answer: B

 

Correct Answer

Answer B is correct because AWS SDK with Amazon SQS for asynchronous inference work is specifically used to decouple callers from long-running or bursty model processing. It places work on a durable queue and lets workers invoke model APIs independently of the original request.

Incorrect Answers

Answer A is incorrect because that requirement aligns with AWS X-Ray distributed tracing, not AWS SDK with Amazon SQS for asynchronous inference work.

Answer C is incorrect because that requirement aligns with API Gateway request validation, not AWS SDK with Amazon SQS for asynchronous inference work.

Answer D is incorrect because that requirement aligns with server-sent events, not AWS SDK with Amazon SQS for asynchronous inference work.

 

Question 27

The enterprise finance automation team has proposed API Gateway request validation for its design. Which requirement best justifies it?

1. Send known request categories to predetermined models using simple deterministic logic
2. Retry transient throttling or service failures without creating a synchronized retry storm
3. Reject malformed API requests before they reach the GenAI backend
4. Deliver generated output incrementally as tokens or chunks become available

Correct Answer: C

 

Correct Answer

Answer C is correct because API Gateway request validation is specifically used to reject malformed API requests before they reach the GenAI backend. It checks required parameters, payload structure, and API-level constraints at the managed gateway.

Incorrect Answers

Answer A is incorrect because that requirement aligns with static model routing, not API Gateway request validation.

Answer B is incorrect because that requirement aligns with AWS SDK exponential backoff, not API Gateway request validation.

Answer D is incorrect because that requirement aligns with Amazon Bedrock streaming response, not API Gateway request validation.

 

Question 28

The risk and controls team includes Amazon Bedrock streaming response in its architecture. What is its primary role?

1. Choose models through multi-step runtime logic based on request content or measured conditions
2. Enforce API-level request-rate limits before model capacity is overwhelmed
3. Maintain a bidirectional real-time connection for interactive AI experiences
4. Deliver generated output incrementally as tokens or chunks become available

Correct Answer: D

 

Correct Answer

Answer D is correct because Amazon Bedrock streaming response is specifically used to deliver generated output incrementally as tokens or chunks become available. It uses a streaming inference interface so users see partial output before the complete response is finished.

Incorrect Answers

Answer A is incorrect because that requirement aligns with Step Functions dynamic model routing, not Amazon Bedrock streaming response.

Answer B is incorrect because that requirement aligns with API Gateway throttling, not Amazon Bedrock streaming response.

Answer C is incorrect because that requirement aligns with WebSocket delivery, not Amazon Bedrock streaming response.

 

Question 29

The enterprise search architecture group plans to adopt WebSocket delivery. Which outcome should drive that decision?

1. Continue serving requests when the preferred foundation model is unavailable or fails
2. Stream one-way incremental model output from server to browser over an HTTP connection
3. Maintain a bidirectional real-time connection for interactive AI experiences
4. Return a model response within the same request-response interaction

Correct Answer: C

 

Correct Answer

Answer C is correct because WebSocket delivery is specifically used to maintain a bidirectional real-time connection for interactive AI experiences. It keeps a persistent connection through which clients and servers can exchange messages asynchronously.

Incorrect Answers

Answer A is incorrect because that requirement aligns with model fallback mechanism, not WebSocket delivery.

Answer B is incorrect because that requirement aligns with server-sent events, not WebSocket delivery.

Answer D is incorrect because that requirement aligns with synchronous Amazon Bedrock invocation, not WebSocket delivery.

 

Question 30

The insurance automation team has proposed server-sent events for its design. Which requirement best justifies it?

1. Stream one-way incremental model output from server to browser over an HTTP connection
2. Decouple callers from long-running or bursty model processing
3. Trace model API calls across application and service boundaries
4. Retry transient throttling or service failures without creating a synchronized retry storm

Correct Answer: A

 

Correct Answer

Answer A is correct because server-sent events is specifically used to stream one-way incremental model output from server to browser over an HTTP connection. It uses an event stream that is simpler than full bidirectional websockets when the client mainly receives updates.

Incorrect Answers

Answer B is incorrect because that requirement aligns with AWS SDK with Amazon SQS for asynchronous inference work, not server-sent events.

Answer C is incorrect because that requirement aligns with AWS X-Ray distributed tracing, not server-sent events.

Answer D is incorrect because that requirement aligns with AWS SDK exponential backoff, not server-sent events.