AWS Certified Generative AI Developer - Professional AIP-C01

A specialty coffee company has a mobile app that uses Amazon Bedrock and a three-stage prompt chain to generate personalized coffee roast profiles. The prompt chain converts user inputs into structured metadata, retrieves relevant coffee-roast logs, and produces a personalized roast recommendation for each customer.

Users across multiple AWS Regions report inconsistent roast recommendations for identical inputs, slow inference in the retrieval step, and unsafe recommendations, such as brewing at excessively high temperatures. The company must improve output stability for repeated inputs, application performance, and the safety of the application's outputs. The updated solution must ensure 99.5% output consistency for identical inputs and inference latency of less than 1 second. The solution must also block unsafe or hallucinated recommendations by using validated safety controls.

Which solution will meet these requirements?