Each morning, a rental car company needs automated insights about the previous day’s reservation demand. Data should stream to S3 in near real time, detect high-demand vehicles per location, and appear on a dashboard that refreshes automatically. Which approach delivers this with the least development effort?

This option is the most efficient because Kinesis Data Firehose directly streams data to S3 with minimal configuration, handling scaling and retries automatically. QuickSight ML Insights provides built-in anomaly detection (high-demand outliers) and visualization capabilities, requiring no custom model development or SageMaker integration. This significantly reduces development effort compared to building and deploying a custom Random Cut Forest (RCF) model in SageMaker. The other options are less optimal: Using Kinesis Data Streams requires more operational overhead and custom code for S3 delivery compared to Firehose. Training and using an RCF model in SageMaker, while effective for anomaly detection, adds significant development and maintenance complexity, which contradicts the "least development effort" requirement. QuickSight ML Insights offers a serverless, managed alternative for this specific task.