Task 2.2: Implement Model Deployment Strategies. AIP-C01 Practice Questions (Page 3)

Questions 11–15

1. 11application · medium

   A financial services company runs a real-time fraud detection system that processes a consistent, high-volume stream of transactions 24/7. The system uses a foundation model in Amazon Bedrock for transaction risk scoring. The compliance team requires predictable latency and no throttling during peak transaction hours. Which deployment strategy should the company choose?

   • AConfigure Amazon Bedrock provisioned throughput with enough model units to handle the peak transaction volume.
   • BDeploy the model to a SageMaker AI endpoint with a single model container and auto scaling.
   • CInvoke the model on demand from an AWS Lambda function for each transaction.
   • DUse a container-based deployment on Amazon EKS with GPU nodes and horizontal pod autoscaling.
   Reveal Solution Discussion Select an answer first
2. 12foundation · easy

   In an API-based model cascading implementation, what typically triggers escalation to a larger model?

   • AThe smaller model is unavailable
   • BThe smaller model indicates low confidence or fails to meet a quality threshold
   • CThe user explicitly requests a larger model
   • DThe larger model is always called first for every request
   Reveal Solution Discussion Select an answer first
4. 13foundation · easy

   What is a primary cost consideration when using AWS Lambda to invoke foundation models on demand?

   • AYou are billed only for the storage of the Lambda function code
   • BYou are billed for the Lambda execution time and the model inference requests
   • CYou are billed a flat monthly fee for each Lambda function
   • DYou are billed only for the provisioned concurrency of the Lambda function
   Reveal Solution Discussion Select an answer first
5. 14foundation · easy

   A developer wants to invoke a foundation model in Amazon Bedrock only when a new file is uploaded to an Amazon S3 bucket. Which AWS service can be used to trigger the invocation automatically?

   • AAWS Lambda function with an S3 event notification
   • BAWS Step Functions state machine with a wait state
   • CAmazon EventBridge scheduled rule that runs every minute
   • DAmazon API Gateway REST API with a custom authorizer
   Reveal Solution Discussion Select an answer first
6. 15foundation · easy

   Which container deployment pattern is commonly used to improve GPU utilization for LLM inference?

   • AUsing a container with a CPU-only instance type
   • BStoring the model in an S3 bucket and mounting it to the container
   • CRunning multiple model replicas behind a load balancer
   • DDeploying a single container with the model and no scaling
   Reveal Solution Discussion Select an answer first