Today
In spring every dealer needs spreader vanes, coulters, nozzles and knives at the same time. The dealer writes the pre-season order from gut feeling or from last year, and spare-parts sales does not know which machines now run in that area. So one dealer holds too much, and another lacks the part exactly when the farmer wants to spread in the weather window. Then the part goes by express, and the machine stands still anyway.
An example
A dealer in the Emsland has 64 ZW 3,000 l fertiliser spreaders from Landrik in its area, 38 of them older than five years. Last spring it pre-ordered 40 spreader vanes and reordered 72 by express. For the new season demi proposes 110 vanes, 24 agitator fingers and 16 metering slides and says why: consumption per machine over the last three seasons, six open warranty cases with worn vanes and nine new spreaders since the autumn. The dealer confirms the pre-season order in the dealer portal in November. Across all 85 dealers, planning sees that 4,200 vanes must be ready by mid-February and tops up the supplier order.
Your data, wherever it lives
Deliveries with serial numbers and parts consumption sit today in the ERP, pre-season orders and dealer stock in the dealer portal, warranty cases in tickets and e-mails, operating hours in telematics, season planning in spare-parts sales' spreadsheets. demi reads, understands and links all of it, with your rights. So the AI calculates a reasoned pre-season order for every dealer in minutes, and your team and the dealer decide what goes into stock.
Where the data lives today
- ERP
- Web & portals
- CRM
- Telematics
- Tickets
- Office files
How demi solves it
Three layers in this order: connect the data first, then the rights, then the AI.
- 1 · ConnectData Layer
The case needs data from the ERP, the dealer portal, CRM, telematics and spreadsheets. demi replaces nothing and links, per dealer, the machines in their area by serial number, type, year and equipment to their BOMs, the parts consumption of past seasons, open warranty cases and, where the farmer shares them, operating hours and hectares from telematics. So every line shows which machines it is meant for.
- 2 · PermitPermission Layer
Each dealer sees only their own machines, their proposal and their terms; telematics data is used only where the farmer has shared it. The proposal orders nothing: spare-parts sales checks it, the dealer confirms the pre-season order in the dealer portal, and only then does it become an order. Who changed what is logged.
- 3 · ActIntelligence Layer
A statistical method calculates the quantities from consumption per machine and season, the number of machines and their age; warranty cases and known wear patterns are added as separate, visible building blocks. The language model only writes the reason for each line, with its source. From all dealer proposals together, planning sees which parts it must order from suppliers or make itself before the season.
How it runs
- 1
demi reads per dealer the machines in their area with serial number, year and equipment.
- 2
Parts consumption of past seasons, open warranty cases and shared operating hours are added.
- 3
The agent calculates a pre-season order per dealer and notes for each line which machines it is for.
- 4
Spare-parts sales checks it, and the dealer confirms in the dealer portal.
- 5
From all pre-season orders, planning sees what must be ordered or made before the season.
What changes
Parts are at the dealer before the season starts
Fewer express shipments and less excess stock
Every pre-season order with a reason and source per line
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
+216 hfreed per year
Assumptions and calculation
demi calculates a reasoned pre-season order per dealer; your team and the dealer approve.
3 h → 2.1 h per case
How we calculate
20 dealer pre-season orders × 3 h × 30% × 1.0 (manual) × 12 months = 216 h
Value per year, before demi costs
Freed per year+216 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
+€14,040
gained per year
216 h ÷ 1,760 h per role = 0.1 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 a language model calculate the quantities?
No. The quantities come from a statistical method on consumption per machine and season; warranty cases and wear patterns are added as separate, visible building blocks. The language model only writes the reason and names the source for every line.
How does demi know which machines run in a dealer's area?
From deliveries with serial numbers in the ERP, the dealers' handover records and the warranty and service cases since. If a machine was sold on used and turns up in service at another dealer, the agent shows it and asks instead of quietly moving it.
Does a dealer see the figures of the others?
No. Each dealer sees only their own machines, their proposal and their terms. The total across all dealers is seen only by spare-parts sales and planning.
What if the season starts earlier or later?
The pre-season order sets what is in stock, not when the weather comes. During the season demi compares sales and pre-season order per dealer and shows early where a part is running short. So spare-parts sales can move stock before a machine stands still.
