Public portfolio showcase. The production/core engineering system and operational project data remain private.
CP Engineering explores how an engineering demand can be transformed into a traceable technical/commercial decision without hiding assumptions, missing information, risk, cost or revision history.
This public repository is a deliberately reduced demonstration built with synthetic engineering data only.
Engineering projects often lose traceability between:
- the original demand;
- requirements and scope boundaries;
- risks and mitigations;
- concept alternatives;
- cost and commercial price;
- revision changes;
- the final human decision.
A dashboard that simply produces a score is not enough. The system must show what is known, what is estimated, what is calculated, what is missing, and what changed.
Demand / Intake
-> Requirements
-> Scope
-> Risks
-> Concept
-> Evaluation
-> Cost
-> Quote
-> Decision Readiness
-> Human Decision Gate
- revision-aware engineering workflow;
- explicit provenance / data nature;
- requirement and scope traceability;
- risk as decision support, not automatic approval;
- separation between internal cost and commercial sale price;
- incomplete quote behavior when required values are unknown;
- decision-readiness signals instead of an opaque GO/NO-GO score;
- fail-closed human approval;
- simple deterministic domain logic with tests.
This repository is not a certified engineering calculation package and does not issue technical approval, legal compliance, safety certification or regulatory sign-off.
The demo is inspired by systematic engineering practice, but it does not reproduce proprietary textbooks, paid technical references, customer drawings, standards, manuals or copyrighted engineering material.
The live demo uses a fictional project:
AX-17 Modular Inspection Fixture
All companies, values, requirements, risks, costs and decisions in the demo are synthetic.
After GitHub Pages deployment:
https://flavio459.github.io/cp-engineering-showcase/
See docs/ARCHITECTURE.md.
See docs/PUBLICATION_BOUNDARY.md.
No build step or paid service is required.
Open index.html through a static HTTP server.
For deterministic logic validation:
node test.jsHTML · CSS · JavaScript · JSON · GitHub Actions
The private core uses a broader TypeScript / Next.js architecture with domain modules, repository abstractions, testing and durable-persistence work. Those implementation details remain separate from this public showcase.
Flávio Souza Barros
Engineering × AI × Automation × Project Systems
Source-visible for portfolio evaluation only. It is not released under an open-source license. See LICENSE.