Transaction throughput over time, sliced by rail_name so you can see which rails are growing or contracting. The stacked bar is the daily-rail trend; the clustered bar is the period total per type.
Count of distinct Posted transfers — multi-leg transfers (a wire + its receiver leg, an ACH batch + its individual credits) count ONCE per transfer_id. The metric leadership asks about. Sibling Transfer Legs KPI exposes the raw matview row count (all statuses, per leg) — the gap is the documented predicate scope, not a defect.
Raw count of all transaction legs in the period — per-leg (not per-transfer), all statuses (incl. failed). Matches the App Info sheet's <prefix>_transactions row_count exactly. Headline pair (this + Total Transactions) makes the documented predicate-scope gap visible.
Total transfer count per active business day. Averaged over days that had any activity; zero-volume days don't surface in the underlying dataset. Rendered as integer because the underlying datum is a count.
Each day's transfer count, stacked by rail_name so rail composition + trend show together. Demo-data caveat: apparent multi-week empty stretches + weekend gaps reflect the bundled demo's short seed window (90 days) + non-business-day cadence — not data outages. Real deploys with full-history ETL won't show them.
Total transfer count over the selected period, per rail_name. Log-scale Y axis — one rail typically dominates an executive period; log scale makes the rest still readable at the same glance.