Today
Order entry retypes purchase orders from PDFs, and every customer uses a different layout. Differences from the quote in price, quantity or date only surface in production or at invoicing.
An example
Talwerk Antriebstechnik receives a purchase order with 14 lines as a PDF. The agent maps the customer's part numbers and finds that line 7 still carries last year's price. It passes that line to the account manager, while the other 13 are ready to be created.
Your data, wherever it lives
Orders arrive in the order inbox as PDF attachments, in the mail body or as fax scans today, some from customer portals, while quotes sit in the CRM, articles and prices in the ERP and customer part numbers often in spreadsheets. demi reads, understands and links them per order line, with your rights. The checked order draft is ready in minutes instead of being retyped, and order entry decides what is created.
Where the data lives today
- E-mails
- PDFs
- Scans & paper
- Web & portals
- CRM
- ERP
- Office files
How demi solves it
Three layers in this order: connect the data first, then the rights, then the AI.
- 1 · ConnectData Layer
Data sources in this case: the order inbox, ERP and CRM. demi replaces nothing and links every order to customer, articles, prices and quote. Customer part numbers are mapped to your own article numbers.
- 2 · PermitPermission Layer
Order entry approves each order before anything is written back, and deviations above a set limit go to the account owner. Every write is recorded.
- 3 · ActIntelligence Layer
The agent reads the lines, maps customer part numbers and compares price, quantity and date line by line with the quote and your data, by fixed rules rather than estimates. It then prepares the order as a draft, each line linked to its place in the purchase order.
How it runs
- 1
The purchase order arrives by e-mail or as a PDF attachment.
- 2
The agent reads lines, quantities, prices and requested dates.
- 3
Customer part numbers are mapped to your own articles.
- 4
Each line is compared with the quote and the ERP; deviations are flagged.
- 5
Order entry confirms, the order goes back after approval and the confirmation goes out.
What changes
No more retyping of purchase orders
Price and quantity differences caught at entry
Faster order confirmations
What does it bring you?
Check it with your own numbers
Every starting value is an assumption you can change. The calculation sits right below, step by step.
Your numbers
Result
+300 hfreed per year
Assumptions and calculation
demi reads the mail and PDF, checks them against quote and master data, and books the order once you approve.
10 min → 5 min per case
How we calculate
300 orders × 10 min × 50% × 1.0 (manual) × 12 months = 18,000 min = 300 h
Value per year, before demi costs
Freed per year+300 h
Assumption: labour cost of €49.50 per hour in German manufacturing in 2025 (Destatis), plus workplace, IT and overheads.
= Value per year, before demi costs
+€19,500
gained per year
300 h ÷ 1,760 h per role = 0.2 full-time roles
An estimate, not a promise. The starting values are assumptions for an industrial company with 150 to 500 people.
Combine with more cases and get the packageQuestions about this use case
Does it work with every customer's order layout?
Yes. The agent needs no template per customer; it reads lines, quantities, prices and dates from different layouts, scans included. For a new customer, order entry checks the first orders more closely. Every value read links to its place in the document.
How are customer part numbers mapped to our articles?
Through the mapping already in your ERP and through earlier orders from the same customer. If a number is new, the agent proposes matching articles with its reasoning. Once you confirm a mapping, it is stored and used next time.
Does demi create the order by itself?
The agent prepares the order as a draft and writes back only after approval. Order entry confirms it, and deviations above a set limit go to the responsible account owner. Every write is recorded with person and time.
What if the data in our ERP is wrong?
The agent shows every conflict between order, quote and master data instead of silently taking one value. If quote and price list disagree, for instance, the responsible person gets a notice. Master data errors become visible and can be fixed at the source.
