In short
- We automate the documents and messages around planning: supplier confirmations, delay notices, advance ship notices, goods receipt, supplier documents and exception lists.
- Forecasting and planning calculations stay in your ERP or planning tool. We use their output and do not sell a replacement.
- Planners and buyers decide on orders, expediting, alternative sources and stock levels. The AI prepares proposals with their reasoning.
- Documents that carry legal responsibility, such as customs declarations, are prepared but submitted by the responsible person.
What we automate, and what we leave to planning software
Demand forecasting, safety stock calculation and network planning are the job of planning software and, in larger companies, data science teams. Your ERP or planning tool already produces numbers. The gap we close is the one between those numbers and the real world: the supplier emails, the confirmations, the delivery notes, the exceptions that nobody has time to read. We read them, compare them with your orders and plan data, and present what needs a decision.
A typical workflow: delay in a supplier confirmation
This is a typical process, not a client project.
- Intake: a supplier sends an order confirmation or a message about a delay, as an email or PDF, in any layout.
- Reading: the AI reads item, quantity, price, delivery date and any reason given, and finds the purchase order.
- Comparison: it compares the confirmation with the order and the required date, and lists every difference with both document lines.
- Impact: it looks up which sales orders or production orders depend on the item and which would miss their date.
- Proposal: it drafts options for the planner, such as accepting the new date, asking for a partial delivery or contacting an alternative supplier, each with the facts behind it.
- Decision: the planner or buyer decides. Replies to the supplier go out after approval.
- Update: the new date is written to the ERP, with a log entry of what was read and who approved.
Other flows follow the same pattern.
Advance ship notices and goods receipt
- Input: advance ship notice, delivery note and goods-received record.
- AI prepares: compares announced and received quantities and items, lists shortages and over-deliveries and drafts the claim.
- Warehouse and buyer check: every difference before a claim is sent.
- Output: matched documents and a short exception list.
Stock exceptions and replenishment proposals
- Input: daily extracts of stock, open orders and the planning tool’s proposals.
- AI prepares: flags negative stock, stock-outs, items without movement and orders without a confirmation, and explains each proposal in a sentence.
- Planner checks: every proposal before an order is placed.
- Output: an exception report and draft orders for approval.
Supplier documents
- Input: certificates, insurance, quality documents and master data forms.
- AI prepares: checks them against your requirements, tracks expiry dates and drafts reminders.
- Buyer checks: whether a supplier may continue to be used.
- Output: a checked supplier file and a list of expiry dates.
What stays with planners and buyers, and what we do not do
- Orders and sources: placing orders, choosing suppliers and agreeing prices stay with planners and buyers. No order is placed automatically beyond limits you set.
- Planning calculations: forecasts and stock policies come from your planning tool.
- Customs and compliance: customs declarations are the declarant’s responsibility. The AI reads the commercial documents and prepares data, and the declarant checks and submits.
- People: we do not build systems that allocate tasks to individual workers or evaluate their performance. These are high-risk under Annex III of the AI Act, see risk categories.
Systems and interfaces
Supply chain teams work with an ERP such as SAP, Microsoft Dynamics 365, Odoo or Sage, often with a planning tool, a transport management system, supplier portals and a great many spreadsheets. We connect through the interfaces these systems offer (API, EDI, export, email or a watched folder) and work alongside them where none exists. The names are for orientation only and do not imply any partnership.
Law and risk
- Customs: the declarant is responsible for the accuracy of the customs declaration under the Union Customs Code, whichever tool prepared the data.
- Product safety: importers and sellers in the EU carry duties under the General Product Safety Regulation, including traceable information on manufacturers. Supplier documents that carry this information are worth checking systematically.
- Data protection: supplier and employee contact data fall under the GDPR: a data processing agreement with every provider (Art. 28) and processing in the EU or UK.
- AI Act: Art. 4 asks for sufficient AI literacy among staff, and Art. 50 requires chatbots to say that they are AI. This overview is general information, not legal advice.
How to measure success
Compare a period before and after on your own orders. Useful figures are the time between a supplier message and the planner’s decision, the share of confirmations compared against the order, the number of late deliveries that were known in advance, the number of exceptions found at goods receipt and the time spent by planners on reading and re-keying. We agree the measures at the start and do not promise figures in advance.
Where this applies
- Manufacturing: confirmations, certificates and goods receipt, see AI in manufacturing.
- Logistics: orders, consignment notes and invoice checks, see logistics AI.
- Retail and ecommerce: supplier documents and product data, see AI for retail and AI for ecommerce.
- Construction: subcontractor and supplier documents, see construction AI.
- Procurement in general: see AI in procurement.