A company has a podcast platform that has thousands of users. The company implemented an algorithm to detect low podcast engagement based on a 10-minute running window of user events such as…

While Amazon MSK is suitable for event ingestion, performing complex transformations with a running window typically requires developing and managing custom applications with Kafka Streams, entailing more operational effort than a fully managed service like Kinesis Data Analytics.

Amazon Kinesis Data Streams is good for ingestion, but merely storing the data in Amazon S3 does not address the real-time transformation requirement of calculating features based on a 10-minute running window for immediate inference.

Amazon Kinesis Data Streams is highly scalable for real-time event ingestion. Amazon Kinesis Data Analytics then provides a fully managed platform to easily perform real-time transformations, aggregations, and windowed computations (like a 10-minute running window) on the streaming data, delivering prepared features for inference with minimal operational overhead.

Similar to A, Amazon MSK offers event ingestion, but processing real-time running windows and transformations would require custom application development and management, which increases operational effort compared to Amazon Kinesis Data Analytics.