"We try it for a couple of weeks and see if it will work out," I answered, when she raised the question of her commute. 16 km, an hour and a half each way, by bus. I used to ride 17 km on my ebike in Zurich, rain, snow or shine, for a decade. In half the time, about 45 minutes. Different country, different mode, similar distance. I thought, how much could it matter. It mattered. Less than a week in, it was the thing that ended the internship. The selection process had been thorough. What broke wasn't the plan. It was the part that started the moment the paycheck did.
The selection process ran on two real gigs. One technical, a UX fix on a booking widget, 24 hours, clear success criteria. The other, plant a sapling. Five teak saplings, planned, approved, executed, documented, over 35 km away. More than double the distance the commute would later ask for. Both were stretched for, willingly. The desire to land the role was visible, almost eager. Then the role landed. Something shifts in that moment. I have seen it in myself too. The version of us trying to get in is not the version that shows up once in.
Time was the input. But the output, the actual thing that mattered, was not the time spent. A commute, a planning call, a wait for approval, none of it looks like output, yet without it nothing else happens. There was a bonus for each gig, yes. But something changed dramatically in the relationship once the employment started. The evaluation itself was feedback in both directions, against the same criteria set at the start. We never got to run a full gig cycle once employment began. That gap needed a process of its own.
Although the internship had a premature death, for many reasons well beyond the role design, I committed to fixing this before my next hire. Mulai was born. I spent the next weeks iterating, gig evaluation at the core, gigs sitting under projects, user roles, a workflow, a timesheet to keep focus. Gigs got qualities too, cadence being one, carrying its own logic for recreation. And then I saw it. Every conversation that had gone well, in the interviews, in the gigs themselves, had one thing in common. The expectation was prepared before the conversation, not during it.
Mulai is live now, mulai.ch, Vidai to Mulai, seed to sprout. What I needed was right there in front of me. A gig with a placer and a doer, a start and an end, and a reflection asked of both, the part every tool I had used somehow skipped. So I built it, gig by gig, each one assigned to myself and to a few agentic friends along the way. Not finished. Not being sold. Built slowly, used on real work, every day. Work, kept honest.
What comes next is more prototypes, more gigs, planted the same way. Mulai becomes the language, with clients and with suppliers, to set expectations before the work begins. To the vendors I am talking to: this is not about what a Swiss address can pay. It is about the jobs we create for young people willing to learn this way. Take that risk with me. If you build full stack, and small business problems excite you more than scale ever could, I would love to talk too. Not for capital. For time, and for partnership. Bring a sapling. We will find the soil.