🧪 Data Science Real Time Analytics Simulation

Real-time analytics means computing meaningful statistics on data that never stops arriving, without ever looking back at the whole history. This simulation drives five synthetic telemetry streams — CPU load, requests/sec, latency, error rate and memory — through a live exponentially weighted moving average and z-score control chart, the same incremental math behind production observability and statistical-process-control alerting. Each incoming event updates its stream's EWMA and variance in constant time, a bar rises and falls with the smoothed level, and a scrolling trail of raw samples behind it is colored by how many standard deviations from baseline it landed. Tune the ingest rate, the smoothing factor α, and the alert threshold, or inject a manual spike, and watch how quickly — and how noisily — the detector reacts.