| Thread | M8×1.0-6H |
| Load | 14.2 kN |
| Ø12.5 H7 | 12.507 |
| Ra 1.6 | Ra 3.2 |
Drawings, reports, supplier files, intranet tools…
Build a source of truth
for your design engineers.
Auto-collect data, auto-flag errors. No-AI repeatable checks.
Built by a mechanical engineer, 10+ years in aerospace.
Book intro callScenarios
How do you detect these?
- The drawing is at Ø45, but the stress report and the assembly tooling document still say Ø42 — outdated data in reports.
- A supplier goes bankrupt; bearing PB-15B is not procurable after 2028. Which assemblies still use it?
- Three suppliers return inspection plans. Supplier 2 attached a Type Q plan where Type B was required.
- An audit finds a banned coating still in use. Which released drawings are affected?
Scenarios
How do you detect these?
Outdated data in reports
The drawing is at Ø45. The stress report and the assembly tooling document still say Ø42.
Non-procurable items
A supplier goes bankrupt — bearing PB-15B is not procurable after 2028. Which assemblies still use it?
Supplier email attachment error
Three suppliers return inspection plans. Supplier 2 attached a Type Q plan where Type B was required.
Audit finding's immediate impact
An audit finds a banned coating still in use. Which released drawings are affected?
These answers already live in your data. The work is getting them into one central source of truth — where every engineer sees the latest status, without opening drawings one by one and copy-pasting.
Many teams told me a specific check they needed was “surely impossible,” until we found a way to automate it. What you see here is a small sample.
Automated
Centralize everything
in an automated database
No more copy-pasting.
Auto-collect from all sources.
Auto-flagging based on rules you define.
- Drawings
- Stress reports
- Supplier e-mails
- Intranet tool
…and whatever else you have
| Drawing | Coating | Thread | Reserve | Supplier | Flag |
|---|---|---|---|---|---|
| DWG-0010 | Chromium | UNF | 1.02 | Qualified | |
| DWG-0011 | – | UNF | 1.58 | Qualified | |
| DWG-0012 | Chromium | M12 | 2.79 | In audit | |
| DWG-0013 | – | UNC | 3.12 | In-house | |
| DWG-0014 | Anodised | M10 | 1.41 | Qualified | |
| DWG-0015 | – | UNF | 2.05 | In-house |
Already familiar
Data in Google Sheets or Excel. Nothing new to learn.
Use what you have
Chain tools you already have — because off-the-shelf software often doesn't fit you 100%.
No AI by default
Engineering checks need reliably repeatable results, which is not AI's strength.
Ready in 2–4 weeks
Not a two-year IT software project.
Pricing
Yours to keep, forever.
Hate subscription plans? Me too.
One-off cost
Typically €__ – __
Pay once, own the full tool.
Price shaped by data sources, checks and complexity.
One-time purchase.
Pay once, own it forever. No licence fee, no recurring subscription cost.
Full code visibility.
You receive the complete tool, including code. Read, understand, modify it yourself — your freedom.
No AI cost.
Repeatable deterministic checks — no hallucination risk, no monthly AI cost. Use AI only where you consciously want it.
Prevented scenarios
Built by the engineer who needed it.
From engineer, to engineer.
Senior mechanical engineer, 10+ years in aerospace design. Metal production as backbone, end-to-end across all design phases. Worked in design offices from a 50-person Dutch startup to a 2,000-person German factory to global corporate — so the solution fits your culture and IT reality, not the other way around.
I always wanted to see, in one view, where a specific fault lives across a whole design program — instead of opening drawings one by one. Thus, as a mechanical engineer who codes, I fully understand your pain — and know how to bridge it.
FAQ
I can't install things on my company PC. Is it still possible?
That's my specialty: enabling automation in strict corporate environments, using minimal (or zero) external tools.
Most solutions are designed around existing tools already allowed by your company IT.
Let's look at your company toolkit and discover the opportunities.
My team's checks involve manufacturing knowledge — coatings, heat treatment, thread specs. Can you follow?
That's my home ground. I've worked for years directly with production: machine operators, suppliers, shop-floor realities, in German.
Cutting, heat treatment, coating, galvanic treatment, procurement, assembly — the detection rules we define together can build on this shared language, instead of me learning your world from zero.
Does our data leave the company?
The default design philosophy: the architecture stays inside your environment — your browser, your company spreadsheet tool, or a local database.
Where an external tool would genuinely serve you better, it's proposed openly and decided together, within your company's policy.
Can our IT review the code?
Yes, and it's encouraged. The final delivery includes a working setup, full source code, documentation, and a one-page summary written for IT.
How do we make sure it works?
Handover includes acceptance testing on your real data, signed off by your team, so the scope boundary is clear.
I also recommend one of your engineers follows the development iterations: they become your internal expert, familiar with the tool from day one.
Bugs within the agreed scope are fixed free of charge during the warranty period.
New check rules or data sources are typically a small follow-up project, or something your internal expert can already implement, having followed the development.
Does the tool contain AI?
By default, no. Technical decisions in mechanical engineering often need deterministic automation on the front line.
Deterministic means: same design rule, same input, same diagnosis. (Example: if a thread text on a drawing is not written in a specific way, it is reliably flagged as a fault.)
Deterministic assessment is not AI's strength. In most cases it's cheaper and more reliable to leave AI out — ineffective AI use adds unpredictability to check processes, with little added value and high running cost.
But there are cases where AI genuinely helps. I'm glad to help your team make an informed decision on what combination works best for your scope.
Is the tool built purely by AI?
Transparency: partially. As the human engineer, I use my 10+ years of engineering background to uncover the root pain point, define the spec, the architecture and the tech-stack choice.
I shape the detection rules with your senior engineers, and test edge cases with my engineering knowledge.
AI is used as a power coding tool, so a big part of the software can be produced and iterated in days.
That's how you get a working tool in weeks instead of years.
All delivered code is human-reviewed. Cross-verification by additional software specialists can be arranged on request.
What does it cost?
Typically €__–__, fixed price, depending on complexity — see The math above.
You'll get a concrete number and scope definition before committing to anything.
Talk to the engineer who builds it.
Specific design errors you need to catch? A manual process everyone hates? Let's take a look.