Today
Complaints arrive by e-mail, phone and portal and are classified inconsistently. Clusters across products, lines or suppliers only get noticed once many cases have piled up.
An example
At Talwerk Antriebstechnik the agent notices 9 complaints about loose connectors within three weeks. All concern geared motors from assembly line 3, built after a tool change. It opens a corrective action for the line manager.
Your data, wherever it lives
Today complaints arrive by e-mail, phone and customer portal, often with photos, while customers sit in the CRM, lots and suppliers in the ERP, production data in the MES and past cases with their actions in the quality system. demi reads, understands and links them per complaint, product and lot, with your rights. The AI sorts every case as it arrives, finds similar past cases and reports clusters by part, line and supplier before many cases pile up, and quality decides on every action.
Where the data lives today
- E-mails
- Phone & voice
- Web & portals
- Photos, video & audio
- CRM
- ERP
- MES
- QM system
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 quality system, CRM, ERP and e-mail. demi replaces nothing and links every complaint to customer, product, lot, production data, supplier and past cases. demi opens corrective actions in the quality system.
- 2 · PermitPermission Layer
Quality sees every case; sales sees the complaints of its own customers. Quality approves every action and gives it an owner, and every status change is recorded.
- 3 · ActIntelligence Layer
The agent classifies and routes complaints, finds similar past cases and flags overdue actions. It counts trends by component, line and supplier from the records, not from estimates, each with the cases behind it.
How it runs
- 1
Read in complaints from every channel
- 2
Classify them the same way and route them to the owner
- 3
Link them to lot, line, supplier and past cases
- 4
Flag clusters and open a corrective action
- 5
Track actions to closure
What changes
Faster handling of every complaint
Clusters found early
Corrective actions closed on time
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
+160 hfreed per year
Assumptions and calculation
demi sorts each complaint, links it to production data and shows clusters; you act.
20 min → 12 min per case
How we calculate
100 complaints × 20 min × 40% × 1.0 (manual) × 12 months = 9,600 min = 160 h
Value per year, before demi costs
Freed per year+160 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
+€10,400
gained per year
160 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
Which channels does it read?
E-mail inboxes, customer portals, forms and the complaints in your quality system or CRM. Phone complaints come in through the note of whoever took the call. Everything becomes one consistent case.
How are clusters detected?
The agent counts cases from the linked records, grouped by component, defect type, line, lot and supplier, and compares them with the usual level. When a group stands out, it reports it with the cases behind it. You always see which complaints were counted.
Can sales see the status for their customer?
Yes, for the customers they look after. Internal root-cause work and supplier data can stay with quality if you set it up that way. Sales can then answer the customer without chasing the quality team first.
What if the ERP data is incomplete?
Then demi shows the missing link, such as a complaint without a lot, and asks the owner for it. The agent never guesses a lot. The more complete the cases become, the sharper the trends get.
