A transportation company has trained an ML model in Amazon SageMaker to predict delivery times based on real-time traffic data. The company needs to deploy the model in a production environment to meet the following requirements: Ensure high availability to provide predictions without interruptions. Achieve low latency for real-time predictions on delivery schedules. Process input data sizes ranging between 1 KB and 5 MB. Handle unpredictable surges in requests during peak traffic hours. Scale dynamically to match fluctuating demand throughout the day. Which of the following represents the best solution for the given scenario?