Today
Sales engineers match duty points against many curves per series and speed by hand. Many inquiries leave out the medium, temperature or inlet pressure. Rounds of questions follow while the customer waits for a quote.
An example
Hartwig & Brunn Pumpenwerk receives an inquiry for 85 m³/h at 42 m head for cooling water at 35 °C. The agent proposes series KP 80 and KP 100 at 2900 rpm, shows the operating point close to best efficiency on the stored curve and asks for the inlet pressure to confirm the NPSH margin.
Your data, wherever it lives
Today inquiries arrive by e-mail, as filled-in data sheets or as lists of duty points, the curves sit in the PLM and in old, partly scanned data sheets, and materials and earlier quotes in the ERP and in folders on the drive. demi reads, understands and links them per series, speed and medium, with your rights. The AI calculates the duty point on the stored curves and proposes pumps with efficiency and NPSH margin in minutes instead of half an hour, and the sales engineer makes the choice.
Where the data lives today
- E-mails
- PDFs
- PLM
- Scans & paper
- ERP
- Folders & drives
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 PLM, ERP and file share. demi replaces nothing and links the curves of every variant, including curves read from old data sheets, with materials, inquiries and earlier quotes. Flow, head, speed and power are held in consistent units.
- 2 · PermitPermission Layer
Price lists and project prices are visible only to those allowed to see them. The sales engineer approves the technical selection before it goes into the quote, and the approval is recorded.
- 3 · ActIntelligence Layer
The agent reads the duty point from the inquiry, asks for missing values and calculates on the stored curve instead of estimating. It proposes pumps with operating point, efficiency and NPSH margin, each with its curve and data sheet as the source.
How it runs
- 1
The inquiry arrives by e-mail, as a data sheet or as a list of several duty points.
- 2
The agent reads duty point and medium and asks the customer for missing values.
- 3
Matching series are calculated on their curves.
- 4
The agent compares them side by side with efficiency and NPSH margin, each with its data sheet as the source.
- 5
The sales engineer reviews and takes the selection into the quote.
What changes
Faster technical selection
Fewer rounds of questions
Consistent sizing across the team
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
+240 hfreed per year
Assumptions and calculation
demi reads duty point and medium and proposes matching series with their curves; you make the choice.
30 min → 18 min per case
How we calculate
100 sizing requests × 30 min × 40% × 1.0 (manual) × 12 months = 14,400 min = 240 h
Value per year, before demi costs
Freed per year+240 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
+$15,600
gained per year
240 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
Can it use our own pump curve database?
Yes. Your curves live in your selection software, the PLM, spreadsheets or old data sheets. How they get to demi, we clarify in a conversation. Curves from PDF data sheets are digitized once and confirmed by an engineer.
Does the AI estimate the operating point?
No. The operating point is calculated on the stored curve. The language model reads the inquiry and explains the result; the numbers come from the calculation.
What if the medium is unusual?
Then the agent does not simply carry on. It flags that viscosity, density or solids content can change the curve, asks for the values and hands the case to the sales engineer.
Does it check seal and material compatibility too?
Yes, if your resistance data is connected. The agent checks medium and temperature against the materials of casing, impeller and seal and shows what the statement is based on. If data is missing, it says so.
