AWS Certified Generative AI Developer - Professional AIP-C01

By Amazon · Question Mode

QuestionQ43

Foundation Model Integration, Data Management, and Compliance

A financial services company uses multiple foundation models (FMs) through Amazon Bedrock for its generative AI (GenAI) applications. To meet a new regulation governing GenAI use with sensitive financial data, the company requires a token-management solution.

The token-management solution must proactively alert when applications near model-specific token limits. It must also handle more than 5,000 requests per minute and retain token-usage metrics to allocate costs among business units.

Which solution will satisfy these requirements?

Explanation

Amazon Bedrock enforces token quotas per model, so token usage must be estimated with the applicable model’s tokenization rules before inference to identify approaching limits. A scalable Lambda layer can perform that pre-request estimation, emit model- and business-unit-specific custom CloudWatch metrics for threshold alarms, and store detailed usage records in DynamoDB for chargeback reporting. Amazon Bedrock Guardrails do not implement token quota policies, dead-letter queues operate after failed requests, and API Gateway usage plans apply request quotas rather than model-specific token calculations.

Learn more

Community Discussion

No comments yet. Be the first to start the discussion!