Supply chain analytics reference builds, implementing the library’s published KPI definitions wherever one exists, running here over frozen, simulated data. No client data, no live warehouse — every figure is checkable against the snapshot named on its build’s page.
Each panel names the decision its build is shaped for — who owns it, whether it is an operational, tactical or strategic review, and the rhythm the instrument is built to be read on. That is design intent over simulated data, not an observed deployment, and the builds stay separate instruments: there is no combined score over them.
Where is inventory protecting service, and where is it just consuming cash?
On-hand units · 365 days
59.7 days of supply · portfolio close
DecisionInventory/replenishment owner · Operational and tactical · Shaped for a weekly read of the position, with policy changes reviewed on their own cycle
Where is operating cash tied up, which cost pools are attributable, and what contribution remains at a compatible grain?
Cash conversion cycle: DIO + DSO − DPO, in days
−16.4 days cash conversion · 2026
DecisionFinance and supply-chain leadership · Tactical and strategic · Shaped for a read per accounting period, once that period's balances and flows are posted
Did we deliver what we promised — and what does “what we promised” even mean?
OTIF by customer against its order lines · 24 customers
OTIF 88.1% · 12 mo · 22.7k order lines
DecisionCustomer-service lead · Operational and tactical · Shaped for a monthly read on the original-commit cohort, with exceptions read between cycles
Which shipments are late right now — and did the carrier miss the promise, or did the promise move?
On-time vs original promise · 3 as-of dates
OTD 70.0% vs original promise · 10 confirmed of 13
DecisionLogistics/transport manager · Operational and tactical · Shaped for a daily read of open shipments at the pinned as-of, with carrier and lane review on the arrival cohort
Can this decision use the selected run’s data — and who owns what is blocking it?
13 evidence conditions × 3 observed runs
66.7% required evidence · latest run
DecisionData steward/domain owner · Operational · Shaped for a read per pipeline run, before a decision uses that run
Snapshot frozen 2026-07
Showing every reference build.
From reference build to your warehouse
The definitions, the pipelines, and the dashboards behind these snapshots are what the practice builds over live data — the foundation first, then the AI on top of it.