A financial institution processes millions of transactions daily. Their existing on-premises system struggles to provide real-time fraud detection and anomaly alerting due to the sheer volume and velocity of incoming data. The institution needs to build a new platform on AWS that can ingest, process, and analyze high-throughput streaming transaction data in real-time to identify suspicious activities immediately. Which combination of AWS services should a Solutions Architect recommend to meet these requirements?
- AAmazon MQ for message queuing, Amazon EC2 instances for processing, and Amazon DynamoDB for storing alerts.
- BAWS DataSync for migrating on-premises data to Amazon Redshift, and Amazon QuickSight for dashboarding.
- CAmazon S3 for data lake storage, AWS Glue for ETL, and Amazon Athena for querying.
- DAmazon Kinesis Data Streams for data ingestion, Amazon Kinesis Data Analytics for real-time processing, and AWS Lambda for alerting.
Show answer & explanationAnswer & explanation
Correct answer: D. Amazon Kinesis Data Streams for data ingestion, Amazon Kinesis Data Analytics for real-time processing, and AWS Lambda for alerting.
Amazon Kinesis Data Streams is ideal for ingesting high-volume, real-time streaming data. Kinesis Data Analytics for Apache Flink can then perform complex real-time processing, such as fraud detection and anomaly identification, on this data. AWS Lambda can be triggered by Kinesis Data Analytics to send immediate alerts, forming a complete real-time fraud detection pipeline.
Why the other options are wrong
- A. While Amazon MQ can ingest data, and DynamoDB can store alerts, using EC2 instances for processing would require significant operational overhead for scaling, management, and implementing real-time analytics logic, which is less efficient than a managed service like Kinesis Data Analytics.
- B. AWS DataSync is for data migration, Redshift is a data warehouse for analytical queries, and QuickSight is for business intelligence. None of these are designed for continuous, real-time streaming fraud detection.
- C. This combination is primarily for batch processing and analytics on historical data in a data lake, not for real-time streaming fraud detection and immediate alerting.
Real-time Streaming Analytics with Kinesis
This pattern combines Amazon Kinesis Data Streams for high-throughput data ingestion, Amazon Kinesis Data Analytics for real-time processing and analysis, and AWS Lambda for immediate actions or alerts based on detected patterns.
- Enables immediate insights and actions on streaming data.
- Fully managed services reduce operational overhead.
- Highly scalable for high-volume, high-velocity data.
Memory trick: Kinesis Keeps Kicking Knowledge.