My creative workflow

Read. Plan. Build. Break it, fix it.

Four steps. No secrets, no methodology I'm trying to trademark. Just the version of this that actually works for me.

01
Learning

Read before you touch.

Before I pick up a single component, I read. Datasheets, forum threads, the occasional YouTube rabbit hole at 1am. Understanding why a component behaves the way it does is worth three failed prototypes. I'd rather spend a day reading than a week debugging something I didn't actually understand.

This is also where I admit what I don't know. Pretending to understand something is how projects die quietly. Asking early is how they survive.

02
Planning

Short, not long.

Then comes the boring, important part: choosing components and sketching how the system should work. Not a full circuit diagram yet — just enough to know what talks to what, what the power budget looks like, and where the interesting problems will show up.

I try to keep this phase short. Planning for too long kills projects. The plan exists to point me in roughly the right direction; the real answers come out of the next step.

03
Building

Breadboard first, always.

This is the fun part. Breadboard first, always. I build the smallest version of the thing that can actually prove it works. Once the core loop is running — receiver reading signal, motor spinning, light turning on — I add features one at a time, testing each before moving on.

Arduino and Discovery boards get most of my builds; the rule is whatever finishes the job fastest. If an off-the-shelf module makes sense, I use it. If I can learn something worthwhile by doing it from scratch, I do that instead.

04
Testing

Not done until it's been broken.

Nothing is done until it's been tested in the conditions it's meant to run in. A circuit that works on my desk is not a circuit that works on a moving RC car at full throttle, or a street light system in actual rain.

I test, I break things, I fix them, I test again. The goal isn't perfection — it's something I can trust. Something I'd be willing to put my name on, knowing someone else is going to rely on it.

That's the whole playbook.

If any of that sounded like the right fit, let's talk.