AI
Infrastructure
Data
Expecting extraordinary results from AI… on minimal resources?
AI is not magic: without quality data, infrastructure and context, its potential stays a promise.
§Logbook — working notes
Blog
Notes on applied AI, enterprise integration and my own products in production. Originally published on LinkedIn.
AI
Enterprise
GDPR
Integrating AI into a company without losing control of its data
The challenge isn't using AI — it's integrating it as an enterprise service, with an internal API as the control layer.
VehicleOS
AI
OBD-II
What if your car could carry part of the intelligence of a workshop's diagnostic tools?
VehicleOS uses AI to go far beyond "code P0401 detected": discovering ECUs, interpreting data and explaining it in plain language.
VehicleOS
PWA
OBD-II
What if your 20-year-old car could have the dashboard of a new one?
That's what I'm building with VehicleOS: real-time data, explained diagnostics and telemetry, all as a PWA.
AI
Agents
Future of work
Developers will no longer write most of the code
What you delegate is the typing. What you don't delegate is judgement: picking the solution that will hold, and verifying that what you accepted actually works.
FABRIK
CNC
Traceability
FABRIK: identify and track every part after nesting
It imports the nesting CSV and the TopSolid STEP into a single viewer: every part with its reference, size and position, linked to its 3D solid.