The correct solution is Set up a Kinesis Data Firehose stream to automatically receive and buffer incoming data, encrypt it, and deliver it to Amazon S3 in Apache Parquet format for later use. Kinesis Data Firehose is a fully managed service designed for low-latency streaming data delivery with minimal operational complexity. It supports encryption at rest and can natively convert data into Apache Parquet or ORC formats before writing it to S3. This eliminates the need for custom code, external processing, or manual conversion, which drastically reduces operational overhead.

The choice Use Kinesis Data Streams for ingestion, then configure AWS Lambda to trigger an Amazon EMR cluster to perform analysis and write results to S3 is incorrect because while it provides flexibility, it introduces more operational burden. EMR clusters need to be provisioned, monitored, and scaled, and Lambda-to-EMR handoffs require custom integration. This is not the path of least resistance for this scenario.

The option Create a Kinesis Data Firehose stream that stores raw data in S3, and configure a Kinesis Data Analytics app to analyze data stored in S3 is flawed because Kinesis Data Analytics is not designed to analyze static data stored in S3. It processes streaming data from Kinesis Data Streams or Firehose, not from objects already stored in S3.

Another option, Ingest data with Kinesis Data Streams and use Lambda to send it to a Kinesis Data Analytics app, which writes output to S3 in Parquet format, is more complex than needed. This introduces multiple services that must be integrated and maintained, which increases the operational footprint.

This question is important because it tests your understanding of serverless streaming data pipelines and your ability to choose services that reduce undifferentiated heavy lifting. A trick to answering questions like this is to look for the solution that minimizes manual intervention, leverages fully managed services, and directly supports the required data transformations (in this case, encryption and Parquet conversion).

What makes this question challenging is the presence of multiple valid-sounding solutions that all involve Kinesis. Recognizing the difference between Kinesis Data Streams and Kinesis Data Firehose is keyFirehose is built specifically for delivery and transformation, whereas Data Streams is more developer-intensive and requires additional downstream consumers for transformation and delivery.

Full AWS Practitioner Certification Question