What is Amazon Kinesis? A plain-English explainer

What streaming data is

Streaming data is data that is generated and needs to be processed continuously, as a constant flow of small events, rather than collected up and processed all at once on a schedule. Our what-is-a-data-pipeline explainer covers moving data through a series of processing steps more generally, and our what-is-big-data explainer covers the broader characteristics of data at scale, of which velocity — how fast data arrives — is one defining trait that streaming is built around.

The contrast is with batch data, which is gathered over a period and then processed together in one run. Both are legitimate ways to handle data; the right one depends on how quickly you need to react to it.

Batch vs streaming, in one line

Batch processing collects data over a period and processes it together at set intervals; streaming processes each piece of data as it arrives, continuously and with minimal delay. Kinesis is built for the streaming side of that distinction.

What Kinesis actually does

Kinesis captures data as it is generated — from applications, devices, servers or other sources — and makes it available for other systems to read and process, typically within seconds of arrival. Producers write records into a stream continuously, and consumers read from that stream to process, analyse or store the data, all while the flow keeps going.

Because the data is available to consumers almost immediately, Kinesis enables use cases where reacting quickly matters: detecting anomalies as they happen, updating a live dashboard, or triggering an action the moment an event occurs, rather than discovering it hours later in a batch report.

Common uses

Kinesis is commonly used to collect and process application and system logs in near real time, to capture clickstream data from websites and apps for live analytics, to ingest telemetry from IoT devices and sensors, and to feed real-time dashboards and alerting systems that need to react as events happen rather than after the fact.

• Application and system log ingestion in near real time
• Clickstream and user-activity analytics
• IoT and sensor telemetry
• Real-time dashboards, alerting and anomaly detection

How this fits alongside other clouds

Real-time data streaming is not an AWS-only concept — every major cloud offers a comparable capability. Microsoft Azure offers Event Hubs for high-throughput event ingestion, and Google Cloud offers Pub/Sub-style messaging for streaming and event-driven data. All of these solve broadly the same problem as Kinesis — capturing and moving continuous streams of data — within their own platform.

Kinesis and pricing

Kinesis is pay-for-use, typically billed based on the volume of data ingested and the capacity provisioned to handle it, rather than a flat subscription. Exact rates change over time, so check the AWS pricing page for current figures rather than relying on a fixed number here.

Where Kinesis appears in certification study

Kinesis is a core topic in the AWS Data Engineer Associate exam, which covers ingesting and processing streaming data as one of its core themes, and it also appears in the Developer Associate and Solutions Architect Associate exams as one of the standard options for building event-driven and real-time architectures.

This article deliberately stays at that introductory level. How Kinesis streams, shards and consumers are actually configured is exam-depth material, and that is what our /revision study library is built to cover.