Microsoft Azure Data Fundamentals DP-900 Practice Question

Streaming (or real-time) processing handles data continuously as it arrives and produces results with very low latency-often seconds or sub-seconds. This allows the system to raise an alert almost immediately after an overheating event occurs. Batch processing, in contrast, waits until enough data is collected to run a scheduled job, leading to much higher latency. The other options describe characteristics that do not meet the near-instant feedback requirement: extensive historical joins increase latency, scheduled jobs are typical of batch pipelines, and fixed daily costs do not address timeliness.