EDI Transaction Health

Release to payment, and what each failure actually costs

What this demo shows

EDI failures are not IT tickets. They are line stops, chargebacks and blocked payments.

Objective

Every supplier relationship runs the same exchange: we send an 862 shipping schedule, they acknowledge with a 997, they send an 856 ASN ahead of the truck, material is received, and they invoice with an 810. Eight rules check that chain end to end. Each failure is tagged with the consequence it causes, and any failure on a part that stops the line inside 24 hours is escalated to line risk.

Try this

  1. Start on Clean day for a healthy baseline, around 74 percent clean.
  2. Switch to VAN outage. Watch functional_ack dominate: suppliers never confirmed our schedule, so they may be building the wrong thing.
  3. Switch to ASN drift. Quantity, label and cum rules light up together, the signature of a partner slowly diverging from us.
  4. Set Inventory to New model launch ramp and re-run. Same EDI faults, far more line risk, because the parts affected have no cover.
  5. Click any row to open its full eight rule trace and the document flow.

What to look for

  • Escalated to line risk, not the clean percentage. A hundred cosmetic failures matter less than one on a part with three hours of cover.
  • Two different line numbers on purpose. Line stop exposure counts every failure that could stop a line. Escalated to line risk counts only those where the part actually stops inside 24 hours. The gap between them is the triage.
  • Each scenario has a distinct rule signature, which is how a real morning gets triaged: the failure mix names the root cause.
  • Certificate expiry fails while everything else still looks fine. That is the point, it is the failure you otherwise notice on a weekend.
  • The same fault mix produces different business impact depending on inventory. EDI health is only meaningful against the clock.

Synthetic, seeded data. Full rule definitions and thresholds: EDI documentation · related: Line-Down Clock, Inbound Re-Planning.

Exchange health

Rule failures by rule

The mix is the diagnosis. One rule dominating is a systemic fault; a flat spread is normal trading noise. Counts individual rule failures, so one exchange breaking three rules contributes three.

Rule failures by consequence

The same failures, grouped by what they cost, which is the language the business escalates in. The bar is rule failures so it reconciles with the panel on the left; the grey number is how many distinct exchanges are affected. They differ where one exchange breaks several rules of the same class.

Trading partners

Worst first, by line risk then failure count. The certificate column is the earliest expiry across that partner's exchanges.

PartnerStandardLinkExchanges CleanFailuresLine risk Cert expires

Exchanges

Worst first. Click a row for the document flow and the full eight rule trace with measured values and limits. Every rule is shown, pass or fail.

ExchangePartnerPartRelease ASN997 inStops at FailedConsequences