An e-commerce checkout pipeline writes order_amount_usd that is normally distributed around a mean o

A medium Pipeline Design interview practice problem on DataDriven. Write and execute real pipeline design code with instant grading.

Domain

Pipeline Design

Difficulty

medium

Problem

An e-commerce checkout pipeline writes order_amount_usd that is normally distributed around a mean of 84 dollars with a standard deviation near 32. One Wednesday the mean drops to 71. No row is invalid; every value is plausible. Row counts are normal, nulls absent, uniqueness preserved, schema intact. The cause turns out to be a coupon-stacking bug applying twice for one customer segment. The section's pattern: distributional checks compare summary statistics (mean, stddev, quantiles, category mix) of the current run against a rolling historical baseline, and fire when the comparison crosses a z-score or percentage-point threshold. Detect the drift by adding a distributional-check transform whose name states the statistic, the baseline window, and the threshold.

Practice This Problem

Solve this Pipeline Design problem with real code execution. DataDriven runs your solution and grades it automatically.