G2

Atharva P.

Cloud BI Engineer

Enterprise (> 1000 emp.)

7/30/2026

"Powerful, Low-Latency Stream Processing with Amazon Managed Service for Apache Flink"

4.5/5

What do you like best about Amazon Kinesis Data Analytics?

We previously used Amazon Kinesis Data Analytics for real-time stream processing. AWS later renamed the service to Amazon Managed Service for Apache Flink, and with the retirement of SQL Applications, we've transitioned our workloads to Apache Flink applications while maintaining the same streaming architecture.

Amazon Kinesis Data Analytics allowed us to build real-time streaming pipelines without managing Apache Flink clusters ourselves. I particularly liked support for event-time processing, stateful streaming, windowed aggregations, checkpointing, and automatic scaling. After migrating to Amazon Managed Service for Apache Flink, these capabilities became even more powerful while maintaining operational simplicity.

The service integrates naturally with Kinesis streams, Kafka, Amazon S3, and CloudWatch. Performance has been excellent for processing millions of streaming events with very low latency. Automatic scaling and managed infrastructure significantly reduced operational overhead compared to self-hosted Flink clusters. Review collected by and hosted on G2.com.

What do you dislike about Amazon Kinesis Data Analytics?

Developing complex streaming applications requires familiarity with Apache Flink concepts, making the learning curve steeper than traditional batch processing. Debugging distributed stream processing jobs can also be challenging. Review collected by and hosted on G2.com.

What problems is Amazon Kinesis Data Analytics solving and how is that benefiting you?

The platform enabled near real-time analytics on streaming application data without requiring us to manage streaming infrastructure.

Example: Application logs and clickstream events were continuously ingested, enriched, aggregated, and written into analytical data stores. Business dashboards reflected user behaviour within seconds instead of waiting for scheduled ETL jobs, allowing operational teams to react much faster. Review collected by and hosted on G2.com.