In short
- Quoting a configurable product combines four different jobs: reading documentation, building the bill of materials, running the calculation and converting formats. Each one consumes the same expert's time.
- Calculation logic usually lives in spreadsheets whose rules only their author knows — that is not a tooling problem, but knowledge that was never written down.
- Preparing a quote ≠ deciding the price. The tool takes over the repetitive part; judging a difficult case and setting the price stay with the person.
- The first step is establishing which stage the quote gets stuck at the longest — that is where shortening the cycle pays off most.
- If nobody can articulate the calculation rules, extracting them is the first stage. A tool preserves logic that exists; it will not invent rules the company never had.
Where the quote gets stuck
Quoting a configurable product is work that combines reading technical documentation, knowing the materials catalog, calculation logic, and a feel for what everything actually costs. Few people can do it, and the ones who can are overloaded.
It starts with analyzing documentation — pulling dimensions, materials, and quantities that determine cost out of 2D drawings, 3D models, and PDFs. Then comes the bill of materials, mapping line items to the plant's catalog, and selecting components and substitutes. The calculation itself usually lives in Excel sheets whose logic only the author understands. Finally, the same quote gets converted into different formats — internal, client-facing offer, tender, and a translated version for export. Every stage eats up the time of an expert who is in short supply.
What can realistically be built
An assistant that reads incoming technical documentation and extracts the parameters that matter for cost — a first draft of the quoting data appears on its own, and the expert reviews and corrects it. A tool that generates a bill of materials from the plant's catalog, proposing substitutes when something is missing. Consolidating calculation rules scattered across spreadsheets into a single tool that learns the decisions of specific quoting engineers. Comparing a new inquiry against the history of past jobs, so earlier calculations can serve as a reference point. Converting a finished quote between formats without retyping it.
The point is to take the repetitive part off the expert's plate, so their time goes toward what nobody else can do for them — judging difficult cases and making the pricing call. The machine doesn't do the quoting instead of them; it prepares the ground.