Today
A supplier declares force majeure, a port is closed, a sub-supplier's plant is on fire. Management wants to know by evening which customers, orders and revenue depend on it and how long stock will last. Purchasing, planning and sales pull bills of materials, stock and order lists from several systems into one spreadsheet, and two days later there is an answer that is already out of date.
An example
On Monday morning the supplier of a surfactant used in 31 cleaner recipes declares force majeure after plant damage, expected to last six weeks. Management at Orvelon asks demi what this means. Four minutes later the answer is there: stock and goods in transit last 19 days, after that 64 orders from 23 customers worth 1.4 million euros in revenue stall, with two dairies and their alkaline cleaners for CIP cleaning hit hardest. A second supplier is approved for 22 of the 31 recipes and, according to its last quote, can deliver 60 % of the volume; the other 9 recipes would need customer approval. Management decides on allocation and the alternative order the same day, and sales informs the customers with released texts.
Your data, wherever it lives
Bills of materials and recipes sit in the ERP and the PLM today, stock, purchase orders and customer orders in the ERP, customer plans and priorities in the CRM, approved alternative suppliers in purchasing spreadsheets, and the notice of the disruption arrives by e-mail or appears on the web. demi reads, understands and links all of it, with your rights. The AI works through a failure in minutes instead of days, and you decide which way out applies and which customer is supplied first.
Where the data lives today
- ERP
- PLM
- CRM
- Office files
- E-mails
- Web & portals
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 ERP, PLM, CRM, purchasing spreadsheets, e-mail and news on the web. demi replaces nothing and links every supplier, through its items, with bills of materials and recipes, stock, open purchase orders, customer orders and customer plans. From every raw material and bought-in part, a path then leads to the customer and the revenue that depends on it.
- 2 · PermitPermission Layer
Management, purchasing, planning and sales see the calculation, each to the extent they see the data today; individual customers' revenue and terms stay with the roles allowed to know them. A scenario changes nothing in the ERP, and people decide which customer is supplied how. Who ran which scenario is logged.
- 3 · ActIntelligence Layer
The agent turns the question into a scenario with supplier, items, start and duration and shows it to you for confirmation. demi calculates stock coverage, affected orders, customers and revenue exactly through bills of materials and dates; the language model only explains the result. The agent also looks for ways out, such as approved alternative suppliers, substitute recipes or stock at other sites, and marks what still needs checking or customer approval.
How it runs
- 1
You ask in plain language, for example: what if supplier X fails for six weeks from Monday?
- 2
The agent turns it into a scenario with items, start and duration and shows it to you for confirmation.
- 3
demi calculates how long stock and goods in transit last and which orders stall from which day.
- 4
You see affected customers, revenue and delivery dates, and next to them the ways out with their conditions.
- 5
You decide on the way out and the allocation, and sales releases what goes to customers.
What changes
An answer to a failure in minutes instead of days
Clarity on which customers and revenue are really affected
Ways out and their conditions at a glance
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
+96 hfreed per year
Assumptions and calculation
demi traces the outage through bills of materials, stock and orders; you decide the way out.
10 h → 6 h per case
How we calculate
2 disruption scenarios × 10 h × 40% × 1.0 (manual) × 12 months = 96 h
Value per year, before demi costs
Freed per year+96 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
+$6,240
gained per year
96 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 does demi know which customers depend on a supplier?
Through the chain that is already in your data: supplier, item, bill of materials or recipe, finished product, customer order. demi links it across the ERP and PLM, also over several levels and sites. Where the chain has a gap, such as an item without a bill of materials, the agent shows the gap instead of estimating it.
How exact is the stock coverage calculation?
As exact as stock, open purchase orders and consumption in the ERP. demi calculates it exactly from these data and shows which values it used. If a stock figure is wrong, you correct it and the calculation runs again in seconds.
Does demi also find alternative suppliers?
demi first shows those already approved for the item, with prices and lead times from their latest quotes. Suppliers you use for similar items are added as proposals to check. You decide whether an alternative supplier fits and whether customers have to agree.
Can I play through such a failure in advance?
Yes. You run failures of your key suppliers before anything happens and see where a single supplier carries a lot of revenue. That leads to decisions such as approving a second supplier or holding more safety stock. demi recalculates saved scenarios every month with fresh data.
