The long version.

I studied risk management, then priced it for a living. For five years at KEMPRO Brokers I traded and priced agricultural and financial exposure: futures, options, hedging.

The work was arithmetic with consequences. A wrong number did not fail a test; it landed on a farmer’s season or a business’s margin.

That taught me to look at the downside first. Before the upside, before the demo, before the slide: what happens if this is wrong, and who pays?

I still start there. It is why I ask what a feature costs when it fails, not only what it earns.

Risk also taught me to write down the assumption. A price is only honest if you can show the inputs. I carry that into software: if a decision cannot be explained, it is not finished.

That is also how I make decisions: calculated risk, a clear downside, and no pretending it belongs to someone else.

I joined Mwabu, a UK edtech startup, as employee number three. We built tablet-based learning experiences and educational games, much of it used across Africa, from a Potchefstroom studio.

Small teams teach you the whole system, because there is no one else to hand it to.

Then I spent seven years on frontend architecture at BMW’s IT Hub. There I learned to deliver inside systems that already exist: other teams, other standards.

Scale is not more code. It is more people who have to agree before anything ships.

Today I lead from the trenches: close enough to the work to know when it is wrong. My job is to connect the business decision to the technical one, make the constraint clear, then get out of the way.

The best delivery I have seen was not heroic. It was a team that knew exactly what it was building, and why.

The default is convenient. It listens, decides, nudges and asks for more data than the task needs. Most products drift there because it is easier than choosing.

I started meLink with one rule underneath: the user leads, the technology follows. No ambient listening, no unsolicited action, no dark patterns. The interface can disappear. The decision cannot.

If a feature wins the demo and loses the person, it does not ship.

We Are Still Human is the same argument out loud. The show began with a position, not a product: presence over prediction, identity over automation, people over the systems using them. The systems are getting smarter and we are getting quieter.

I am comfortable being early, and I would rather find out in the build than in a meeting.

Then I wrote the six rules down, because a principle with no implementation is a wish. That is the refusal: not a slogan, a specification.