Stream Processing
Computing over data continuously as it arrives, event by event, rather than in scheduled batches.
Stream processing handles data as an unbounded, continuous flow, computing results incrementally as each event arrives instead of waiting to collect a full dataset. Frameworks like Flink, Kafka Streams, and Spark Structured Streaming maintain running state, windows over time, and handle late or out-of-order events.
Worked example: a fraud system scores each card transaction the instant it happens against a rolling five-minute window of that card’s activity, flagging anomalies in milliseconds instead of in a nightly report. Gotcha: time is the hard part — events arrive late and out of order, so you must choose between event time (when it happened) and processing time (when you saw it), and windowing on the wrong one silently drops or misattributes data.