I'm Guillem.

For about a decade I helped build B2B SaaS used every month by 50,000 employees and students at KLM (the Dutch airline), NS (the Dutch national railway), and BBVA (one of Spain's largest banks), then satellite aerospace software after that. The kind of software where a bug means a payment does not clear, or a passenger reads "delayed" on a platform in Utrecht, or a satellite logs the wrong telemetry on its way over the Atlantic. Those organisations have engineering teams, procurement departments and 18-month implementation cycles, and after enough of those cycles you start to notice that most of the calendar is meetings about the work, not the work.

It started in December 2025, in a conversation with friends about what AI had just made possible in software. One of them put me in touch with a delivery company running 30 vehicles that wanted to digitise a process they were still doing by hand. In 2 days, they had a working prototype.

That speed is why Engranatge exists. At an airline, the same problem would have been an 18-month cycle, most of it meetings and paperwork. A delivery company with 30 vehicles has none of that, so the work is just the work. Same engineering, without the wait. I'm curious by nature and I like building things that take real work off people's hands, so I kept building them.

What I'm for, what I'm against

Three honest things

Three honest things before you decide. I am a developer, not a salesperson, and this page is the longest thing I will ever write about myself. I am in pilot phase, so I have a handful of real cases instead of fifty polished logos, and every one of them is a small business I can name in a private conversation. Given the choice between under-promising and over-building, or the other way round, I will pick the first one every time.

The cause

Machines do the low-value work. Humans do the high-value work.