AWS Certified Data Engineer Associate DEA-C01 Practice Question

DynamoDB is purpose-built for key-value access with single-digit millisecond latency, so using patient_id as the partition key lets the application fetch the latest record for one patient well under the 10 ms target. Enabling a sort key on timestamp keeps recent items together and allows simple queries such as "ORDER BY timestamp DESC LIMIT 1". DynamoDB Streams can capture every insert and invoke an AWS Lambda function or Kinesis Data Firehose to batch the changes into Amazon S3. Converting the batches to columnar Parquet files keeps storage inexpensive and gives good scan performance. Amazon Athena is serverless and can directly run SQL across the S3 data without provisioning or managing infrastructure, satisfying the overnight analytics requirement at very low cost. The other options either cannot meet the latency goal (Amazon Redshift, Amazon RDS), or require continuously running an in-memory cache plus manual snapshot exports (ElastiCache for Redis), which increases both cost and operational complexity.