Skip to content
In progress2025 — PresentDesigner & EngineerIndustrial Machinery, Engineered-to-Order

CPQ Engine

Engineered-to-order industrial machinery quotes in minutes, not days — driven by a custom formula compiler.

Quote – Header & Technical Summary
https://cpq-demo.facundocosimo.eu
CPQ Platform Quote Header and Technical Summary
Open
−0%
Time to quote
Days → 0h
Price-list propagation
0%+
Formula cache hit rate
0
Engineering parameters
Why this exists

Quoting custom industrial boilers lived in a chain of linked Excel files. Pricing formulas were spread across tabs nobody fully understood, dependencies broke silently when a price changed, and a complete quote took four to eight hours of cross-referencing power output, efficiency, pressure, humidity scenarios, and parts assembly. The deep knowledge sat with two senior engineers — when either was unavailable, the entire quoting line stalled.

What it is

At the centre is a custom pricing engine that turns plain-language rules into fast calculations. When a parameter changes, the system figures out exactly which formulas depend on it and recomputes only those — not the whole quote. Changes feel instant instead of grinding through hundreds of cells.

A unified calculator runs 56 engineering formulas (chamber volume, flue gas flow, fan power, biomass consumption, and more) in the correct sequence, and supports best-case and worst-case humidity scenarios in a single pass. The bill of materials is hierarchical and supports configurable parts. Pricing tables can be looked up in four ways — exact match, linear interpolation, upper bound, lower bound — covering every common pricing pattern.

Quote assembly time dropped from four-to-eight hours to fifteen-to-thirty minutes. Price-list updates flow into every open quote within hours, instead of requiring manual rework. Every price-list version and every change is preserved — the full pricing trail is auditable end to end.

Stack

Built with

A pragmatic stack — selected to ship fast without trading away long-term operability.

28 technologies · 5 layers

Frontend

10 tools

  • Next.js 15
  • React 19
  • TypeScript 5.9
  • Tailwind CSS 4
  • Zustand
  • React Hook Form
  • Zod
  • Radix UI
  • Recharts
  • Monaco Editor

Backend

7 tools

  • Next.js Server Actions
  • Prisma 7.4 ORM
  • pg driver
  • server-only
  • docx + Carbone
  • xlsx
  • date-fns

Database

3 tools

  • PostgreSQL
  • Versioned price sheets
  • Natural-key entities

Auth & Security

5 tools

  • NextAuth.js
  • bcryptjs
  • Role-based pricing visibility
  • Approval workflows
  • Audit log

Infra & DX

3 tools

  • Node.js 20+
  • Turbopack
  • Vercel
Highlights

What makes it work

01

Smart recalculation

When a parameter changes, the engine recomputes only the formulas that depend on it — typically 10–20 items out of 200 — instead of grinding through the whole quote. Results feel instant.

02

Best-case and worst-case in one pass

56 engineering parameters run in the right sequence across best-case and worst-case humidity scenarios. Boiler-type overrides and calibration coefficients were carried over directly from the source spreadsheets — the new engine matches the original calculations exactly.

03

Flexible pricing lookups

Pricing tables can be searched in four ways — exact match, linear interpolation, upper bound, lower bound — covering every realistic pricing pattern. Formulas can reference parameters, sibling components, and cross-item costs cleanly, with no copy-pasting numbers between sheets.

04

Versioned Word export

Quotes are rendered to Word documents from versioned templates with the right offer sections included per quote. Each generation is stamped for traceability — old quotes can be reproduced exactly as they were originally sent.