AIP-C01 Operational Efficiency And Optimization Practice Test 2 — 15 Questions

Question 1

When configuring an Amazon Bedrock Knowledge Base, which setting helps optimize the retrieval process for RAG applications?

Accepted Answer

Adjusting the chunk size and overlap. Selecting the appropriate embedding model and defining chunks size are key to optimizing retrieval quality and performance.

Answer

Increasing the number of input tokens

Answer

Deleting the data source

Answer

Disabling encryption

Question 2

You are troubleshooting a SageMaker inference endpoint and need to see the container logs. Which service stores these logs?

Accepted Answer

Amazon CloudWatch Logs. SageMaker inference endpoints run containers on managed EC2 instances, and all stdout/stderr output from those containers is streamed to Amazon CloudWatch Logs under log groups named /aws/sagemaker/Endpoints/<endpoint-name>. This is where you inspect model loading errors, inference failures, and framework-level messages. CloudWatch is the native logging destination for nearly all AWS managed compute services, making it the correct answer here.

Answer

Amazon S3

Answer

Amazon DynamoDB

Answer

AWS CloudTrail

Question 3

You notice that your SageMaker model endpoint experiences cold starts. Which feature can mitigate this?

Accepted Answer

Enabling Provisioned Concurrency. Provisioned Concurrency maintains a set number of initialized instances, eliminating cold starts for serverless inference.

Answer

Increasing the instance count

Answer

Changing the container image

Answer

Using a smaller model

Question 4

Your team uses SageMaker Pipelines for training. How can you optimize the storage costs of the artifacts generated by these runs?

Accepted Answer

Configuring S3 Lifecycle policies. Implementing an S3 Lifecycle policy automatically moves or deletes artifacts after a certain period, keeping storage costs controlled.

Answer

Disabling versioning on the S3 bucket

Answer

Using smaller instance types

Answer

Deleting logs every day

Question 5

Which AWS service is used to centralize and visualize costs for your various SageMaker projects?

Accepted Answer

AWS Cost Explorer. AWS Cost Explorer allows you to view, analyze, and visualize your AWS spending, including SageMaker usage.

Answer

CloudWatch Dashboards

Answer

SageMaker Studio Dashboard

Answer

AWS Trusted Advisor

Question 6

You are using Amazon Bedrock and need to ensure your application respects regional service quotas. Which service allows you to request a quota increase?

Accepted Answer

AWS Service Quotas. AWS Service Quotas is the centralized service for managing and requesting increases for AWS service limits.

Answer

AWS IAM

Answer

Amazon CloudWatch

Answer

AWS Support Center

Question 7

Which TWO of the following strategies are best for reducing the cost of SageMaker training jobs?

Accepted Answer

Using Managed Spot Training. Using Managed Spot Training and instance-appropriate types are standard methods for reducing training costs.

Answer

Keeping training clusters running

Answer

Increasing the number of instances

Answer

Enabling detailed logging for every epoch

Question 8

Which TWO metrics are critical for monitoring an Amazon SageMaker inference endpoint to ensure operational health?

Accepted Answer

Invocations. Option A (Invocations) is correct because it is a core CloudWatch metric for SageMaker endpoints, counting the number of requests sent to the endpoint's InvokeEndpoint API, which directly reflects traffic volume and whether the endpoint is being used as expected. Option D (ModelLatency) is correct because it measures the interval between when the endpoint receives a request and when it returns the response, making it a key indicator of inference performance and operational health. Together, Invocations and ModelLatency let operators detect traffic anomalies and latency degradation that signal endpoint problems. Option B (NetworkPacketLoss) is not a standard SageMaker endpoint metric and is not surfaced for monitoring inference health. Option C (S3StorageUsage) relates to S3 bucket capacity, not endpoint operation, and option E (IAMUserCount) is an identity metric unrelated to endpoint performance.

Answer

NetworkPacketLoss

Answer

S3StorageUsage

Answer

IAMUserCount

Question 9

Which TWO actions should you perform to identify cost savings in an existing SageMaker environment?

Accepted Answer

Review AWS Trusted Advisor recommendations. Analyzing Cost Explorer and using Trusted Advisor are the correct ways to find optimization opportunities.

Answer

Delete all data in S3

Answer

Increase the number of IAM roles

Answer

Deploy models to every region

Question 10

Which TWO approaches help optimize the costs of using Amazon Bedrock?

Accepted Answer

Purchasing Provisioned Throughput for predictable workloads. Option C is correct because purchasing Provisioned Throughput for predictable workloads lets you reserve model units at a fixed hourly rate, which is more cost-effective than on-demand pricing when usage is steady and high, and it also provides consistent throughput. Option D is correct because selecting a smaller, more efficient model for simple tasks reduces the per-token cost, since Amazon Bedrock charges based on the model used and smaller models are significantly cheaper than larger ones for tasks that don't require advanced capabilities. Option A is incorrect because using the most expensive model for all tasks unnecessarily increases costs when cheaper models could suffice. Option B is incorrect because disabling all logging does not reduce Bedrock inference costs and may violate compliance or operational requirements. Option E is incorrect because running every request as a separate job adds overhead and does not optimize costs compared to batching or using appropriate model selection.

Answer

Using the most expensive model for all tasks

Answer

Disabling all logging

Answer

Running every request as a separate job

Question 11

Which THREE factors should you consider when tuning your inference endpoint for cost-efficiency?

Accepted Answer

Auto Scaling policies based on custom metrics. Instance type, auto-scaling configuration, and model optimization (like quantization) are critical levers for cost efficiency.

Answer

CloudTrail log retention period

Answer

Number of S3 buckets used

Question 12

Which THREE metrics are best used for setting up auto-scaling policies on a SageMaker endpoint?

Accepted Answer

InvocationsPerInstance. InvocationsPerInstance, CPUUtilization, and GPUUtilization are the standard metrics to trigger scaling actions.

Answer

IAMUserLoginCount

Answer

S3RequestLatency

Question 13

You are optimizing the prompt handling of a production application built on Amazon Bedrock. The application invokes Anthropic Claude 3 Sonnet with a large, static system prompt that instructs the model to act as a customer support agent for a specific product catalog. Each user request also includes a small variable user message. You want to reduce the per-token cost and latency of these invocations without changing the model or the instructions. Which Amazon Bedrock feature should you use?

Accepted Answer

Enable prompt caching for the static system prompt in the Amazon Bedrock invocation.. Prompt caching is the Amazon Bedrock capability specifically designed to reuse static prompt prefixes, reducing both input token charges and latency for repeated invocations. Because the scenario requires keeping the same model and instructions while lowering cost and latency, caching the static system prompt is the only option that directly addresses the requirement without altering the model or workflow.

Answer

Use Amazon Bedrock Guardrails to filter the system prompt before sending it to the model.

Answer

Use Amazon Bedrock model customization to fine-tune the model with the system prompt as training data.

Answer

Switch to the Amazon Bedrock Converse API and include the system prompt in each request.

Question 14

A generative AI team runs nightly batch inference on Amazon Bedrock for 40 million prompts. Each job currently invokes the model once per prompt with a synchronous Converse call and the job takes nine hours, exceeding the six-hour window. The prompts are independent and the team needs the same model quality. Which change most directly reduces wall-clock time while preserving per-prompt output quality?

Accepted Answer

Submit the prompts through the Bedrock batch inference job API and poll for job completion.. Batch inference is purpose-built for high-volume, latency-tolerant generative AI workloads on Bedrock. It ingests a JSONL manifest from Amazon S3, executes the same foundation model asynchronously, and returns outputs to S3, so quality is unchanged while throughput scales far beyond what a synchronous per-prompt loop can achieve. It also avoids client-side retry storms and quota contention that plague naive looping.

Answer

Switch to a smaller, cheaper model for the entire batch and accept lower quality.

Answer

Increase the InvokeModel requests-per-minute quota and keep the same synchronous loop.

Answer

Move the loop to an AWS Lambda function with a 15-minute timeout and higher memory.

Question 15

A generative AI application uses Amazon Bedrock to generate product descriptions. The application invokes the model for each product individually, resulting in many small requests that are throttled by the model's per-account quota. You need to increase throughput and reduce the number of API calls while keeping the same model and output quality. Which approach is MOST operationally efficient?

Accepted Answer

Implement request batching by grouping multiple product descriptions into a single Amazon Bedrock invocation using a structured prompt.. Batching multiple product descriptions into a single Amazon Bedrock invocation reduces the number of API calls, which directly mitigates quota throttling and improves throughput. It preserves the same model and output quality because the model is unchanged; only the prompt structure changes. This is the most operationally efficient way to handle many small generation tasks under per-account quotas.

Answer

Use Amazon Bedrock provisioned throughput to reserve dedicated capacity for the model.

Answer

Increase the temperature setting to generate multiple variations in a single call.

Answer

Switch to a smaller, faster model to handle the requests more quickly.