Scheduled Legacy Pipeline Atlas

Pipeline × function. Fixed columns, so a lit cell always means the same thing — the near-universal core vs the sparse discriminating tail is legible down each column. Rows grouped by product; the header stays pinned as you scroll.

Each customer's footprint — its product mix and every scheduled pipeline it runs, with what each produces. Cards hide when a filter empties them.

Every pipeline by how often it runs (x, log scale — monthly → every-minute) and how many steps it runs (y, tree size). Real-time Flow loops cluster low-left-ish and light; the heavy overnight RevFind batches sit top-right. Colored by product.

Most pipelines are daily, so they pile onto one column above. Here's just the daily cohort re-plotted by fire time (x, midnight → midnight) × complexity — the overnight batch window (shaded) and the working-day cascade. Same filters apply.

A single day on worker-1. Loops render as frequency-ticked bands; daily/monthly jobs cascade by fire time. Responds to the filters above.