Catch delivery slips before the line stops

The manufacturer builds special-purpose machinery and components and monitors hundreds of open order lines every day across several hundred suppliers. A large share of those are themselves mid-sized firms without EDI — they send order confirmations by email as a PDF, as free text, or not at all.

With a few hundred incoming replies per week, most were processed only superficially — delays surfaced only once production was waiting on the material.

Order numbers arrived in five different notations (“BE-2024-0815,” “240815,” “815/24,” the supplier’s own confirmation number, or no reference at all). The mental matching cost a minute per email, with three system switches per case.

Slips sat unstructured in free text (“Line 4 slips by one week”). There was no analysis of individual suppliers’ delivery reliability — end-customer penalties were the result.

Piko reads the procurement inbox via IMAP and connects read-only to the ERP’s order tables. Every reply is classified (e.g. ‘confirmation with partial slip’) and matched to the right order lines — whether PDF, free text, or EDI.

Deviations are assessed against the ERP target and against downstream demand: a slip with no waiting production order is non-critical; the same slip with an end-customer penalty is escalated immediately — with a source reference (message ID + quoted span).

The ERP remains the system of record; entries are made only after the buyer confirms. Simple confirmations can be configured to pass through automatically.

Processing a supplier reply dropped from around eight to about one minute — at a hundred replies a day, that eliminates a full-time processing role.

Critical slips are caught while there’s still time to react — not when assembly is about to begin.

Delivery reliability per supplier, frequently slipped materials, and average delay in days are a standard query for the first time.