Today
Late orders are often noticed only when the delivery date has passed or the customer calls. By then there is no time left to react. The reasons are scattered across the ERP, the MES and supplier e-mails.
An example
At Talwerk Antriebstechnik the agent warns that order 45-4120 will miss its date by four days because the paint shop is booked out. It suggests an external painter Talwerk has used twice before and shows that the date holds with that option.
Your data, wherever it lives
Customer orders, bought-in parts and material sit today in the ERP, progress and capacity in the MES, supplier dates as PDF confirmations in e-mails, and detailed scheduling often in spreadsheets. demi reads, understands and links them per customer order, with your rights. Every day the AI checks all open orders in minutes, calculates the likely finish date and names cause and countermeasure before the customer calls, and the planner decides.
Where the data lives today
- ERP
- MES
- PDFs
- E-mails
- 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, MES and e-mail. demi replaces nothing and ties every customer order to its production orders, bought-in parts, material, capacity and supplier confirmations. demi writes a newly decided date back only after approval.
- 2 · PermitPermission Layer
Planning and sales see warnings for their own orders. The planner decides on countermeasures, sales approves the message to the customer, and both are recorded.
- 3 · ActIntelligence Layer
The agent projects the completion date from recorded progress, remaining work per routing and capacity, not by estimate. It flags orders at risk with cause and bottleneck and proposes options such as resequencing, extra shifts or subcontracting.
How it runs
- 1
Every day demi checks open orders against progress and remaining work.
- 2
The agent projects the likely completion date.
- 3
If a date is at risk, it names the cause and the bottleneck.
- 4
It proposes countermeasures and shows the effect on other orders.
- 5
The planner decides, and sales informs the customer with a ready draft.
What changes
Fewer late deliveries
Time to react
Customers informed early
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
+108 hfreed per year
Assumptions and calculation
demi flags orders whose date is slipping, with reason and countermeasure; you decide and inform.
30 min → 21 min per case
How we calculate
60 at-risk orders × 30 min × 30% × 1.0 (manual) × 12 months = 6,480 min = 108 h
Value per year, before demi costs
Freed per year+108 h
Assumption: labor cost of $48.62 per hour in US manufacturing in June 2026 (BLS), plus workplace, IT and overhead.
= Value per year, before demi costs
+$7,020
gained per year
108 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
How early does it warn?
As soon as the projection shows the date will not hold. How early that is depends on how current your MES confirmations are. The check runs daily or more often, as you decide.
What countermeasures can it suggest?
Resequencing, extra shifts, an alternative machine, subcontracting or a partial delivery. The agent only suggests what exists in your data, such as suppliers you have worked with. For each option it shows the effect on other orders.
Does it inform sales automatically?
Sales sees the warning for its orders right away. The agent drafts a message for the customer with the new date. It goes out only once sales approves it.
What if the MES confirmations are not up to date?
Then the projection is uncertain too, and the agent says so. Orders with no confirmations for a long time are marked “progress unclear” instead of being shown as on time.
