Software for engineering teams, built by an engineer.
We automate what eats your technical team's hours: running and organising simulations, processing test data, producing reports and controlling documents. Python, data and AI applied with engineering rigour — and verified against your own results.
Scoped projects, with a fixed price agreed before we start.
Expensive engineers doing spreadsheet work.
- ×
Dozens of simulation cases set up, run and checked one by one.
- ×
Results scattered across folders, spreadsheets and emails, with no record of which version produced what.
- ×
Test data processed by scripts only one person on the team understands.
- ×
Simulations too expensive to explore the design space properly.
- ×
Technical reports assembled by hand after every test or revision.
- ×
Documents, revisions and approvals with no traceability of who decided what.
From hand-run cases to automatic results.
We work alongside your team, with the tools it already uses. The engineering knowledge stays with you; we take care of the software around it.
01Simulation automation
Pre- and post-processing, batch case generation, execution and result consolidation in Python — integrated with the solvers your team already uses.
02Test and sensor data
Pipelines that read acquisition files, clean and validate them, and turn them into reliable datasets, plots and reports.
03Surrogate and reduced-order models
Data-driven models — from sparse regression to neural networks — that reproduce expensive simulations in seconds, with the error measured.
04Internal engineering tools
Document and revision control, approval workflows, requirements traceability and technical records — with an audit trail.
05Technical dashboards and reports
Simulation and test results in web dashboards and automatically generated reports, in your company's format.
06Platform integration
CAD, CAE, PLM and AI-for-engineering platforms connected to your data and workflows, without manual rework.
Aeronautical engineering and software engineering.
nolForge's founder is an aeronautical engineer from the Federal University of Uberlândia (Brazil). His undergraduate thesis built data-driven reduced-order models of a two-degree-of-freedom aeroelastic system from CFD/FSI simulations — comparing SINDy with deep autoencoders. Since then he has built production software: systems, automation, cloud and AI.
- Physics before codeUnderstand the phenomenon and the data before choosing the model.
- Models that explain themselvesWe prefer interpretable equations when accuracy is equivalent.
- Validated against dataEvery model is compared with the reference simulation or test.
In the thesis, SINDy reached 0.99 accuracy on the aeroelastic case and reproduced the system response in about 144 seconds on commodity hardware. The autoencoder was slightly more accurate (0.9992) — and was still rejected, because it yields no equations you can inspect.
Small, verifiable and no lock-in.
Fixed scope and price
You know the cost and the timeline before we start. The first project can be a small pilot.
Validated against your data
A delivery only counts when the result matches the reference simulation or test.
Code and documentation are yours
Your team can maintain and extend it without us.
Confidentiality
We work in your environment, with access you control. Happy to sign an NDA.
Remote, worldwide
Time zone compatible with the Americas and Europe.
Engineer to engineer
No middlemen: you talk directly to the person writing the code.
Which technical task would your team never want to do by hand again?
Tell us in a few lines. We reply within one business day.