I had been chatting back and forth with one of two engineers that I really trust on matters that are important on anything technical. Both are what I would call a 'thinking man's engineer'. What I mean by that is that there are engineers who think like business owners, and engineers who don't. I am not sure what it is about software developers, but most of the time, they just don't seem to get it and it has nothing to do with intelligence. It's a different way of thinking, or maybe a different skillset. Engineers that can think with a more creative approach, or have interest in marketing, not sure what it is, but I can tell you it's rare. In my twenty years of working in the software services and SaaS space I have come across it maybe four times. One of these has taken a cushy 9 to 5 job where he will probably be getting paid $250,000 pa working 5 hours per day. I say 5 hours because he is one of those engineers that can be so productive using AI that he will probably figure out a way to code while speaking into his iPhone and dropping his boys off to pre-school. Maybe that will be all engineers in a few years.
The other one is this guy that I have about the same amount of respect for.
I wrote him a one liner on whatsapp if he wants to be CTO of TestVentures. He writes back: "yeah, I'm fine with that." No offer letter. No recruiter. No back-and-forth on title. Just that one line.
He's my first human hire. Until this week, TestVentures was me, three or four AI models running in different tabs, and a long list of clients I wanted to take on but couldn't. Adding a human is a different decision. Adding the right human is the only decision I'd make at this stage. Jan is the right human.
He's not going to be full time, because I don't have the funds coming in, but he is going to be paid on an as needed basis, and the spot is permanently there for him when the revenues are higher. I've built an agency up past 7 figures in less than 12 months before. This will be more of a lifestyle company that also pays well, I am optimistic I can do the build up again.
How we met: a microconf mastermind
We met three years ago in a microconf mastermind.
Microconf is Rob Walling's community for bootstrappers, the people building software businesses without venture money, mostly solo, mostly in the one-to-ten employee band. A mastermind is six to eight of those people who meet weekly on Zoom for a year and pull each other's businesses apart. Real revenue, real customer problems, real decisions. The barrier to entry is a working business and the willingness to put your numbers on the table. Usually there is no cost for these, but in some cases you can pay. I've done both, the free ones are better. Perhaps because with the free ones, you are there by your own commitment and people really want to help you, because the times when you do need help, it's reciprocal and you want frank feedback, even if it is at times too frank.
The conversations are sharper than any incubator I've sat in. Nobody is pitching anything. Nobody needs to look smart for a partner. Everyone is shipping or trying to ship, and the feedback is meant to be raw, open, honest. We are not doing any help to the founders if we don't call them out, or at least tell them exactly what we think.
In our group Jan was one of two engineers. Most of the rest of us were operators or marketers. He had a day job as a senior engineer at a European product company and was building side projects on his own time. He'd already shipped two of them: a developer-tools product that found a few hundred paying users but didn't take off, and a workflow tool that never quite caught. Most masterminds I've been in have at least one person who shows up to absorb advice and never ships. Jan was on the other end. He shipped, broke things, came back the next week with what he'd learned.
About 18 months in, we started building together. A side project, mine, that needed someone who could actually write code. We worked on it for 12 months. Both of us working on something a few hours per day, sometimes a Saturday morning. It didn't become a business. It taught us how to work together. That's the part I cared about.
The thing I can't do: read code
I can't read code. Not really. I can scan a pull request and tell you if it's commented and structured. I cannot tell you if it's good code or bad code, idiomatic Kotlin or hack Kotlin, well-designed or fragile. I'm a 5 out of 10 generalist across 15 areas of business, as I have stated many times. I'm a 2 out of 10 at reading what an engineer wrote. So if I'm hiring an engineer, I can't evaluate the one thing the engineer is paid to do, so I don't do that part. I ask someone I trust to check their code, and I go by feel on how the code looks, acts and works. Their comments when building. It's all by vibe, remember? Vibe living.
What I can evaluate is how someone works. How they think. How they behave when the deal isn't fun yet. How they handle the conversation about equity, or credit, or what to do when the project goes sideways. That stuff I can read. I've been running businesses for twenty years, mostly with one or two other equity partners, and the pattern repeats. The people who're great in the first few months are not always the people you want in month nine and beyond. The ones I'd hire again are the ones who hold up under the boring parts or parts when there are no revenues but they're still showing up. How they handle tough conversations, or who will show up on a Saturday night if there's a problem because they actually care about their work.
Twelve months of working with Jan, on every behaviour I watched, I'd give him about a 9 out of 10. Creativity, business acumen, clarity in the hard conversations, and most of all, thoughtfulness from an owner's perspective.
Most engineers think like engineers. Jan thinks like an owner. He asks who's paying for this, who's going to maintain it, what happens if it doesn't work. The architectural decisions follow from those questions, not the other way around. I've worked with maybe four engineers in twenty years who do that. The rest are great at the code and indifferent to the business. Both kinds are valuable. Only one of them is hireable as a CTO of a two-person operation.
High agency is the trait I'm hiring on
High agency. That's the trait I'd put at the top of the list, the one I think gets the least credit in 2026.
Top companies right now are hiring failed entrepreneurs on purpose. Y Combinator's recent posts on what they look for in early engineers describe it without naming it: someone who has tried, shipped, hit a wall, and kept going. OpenAI and Anthropic are scouting for the same trait in their early-employee hiring. Call it agency, call it ownership, call it "this person does not need to be told what to do next." Whatever you call it, in an era where AI handles a growing slice of the actual work, the person who can decide what to do is rare.
Jan hasn't had an exit yet. He's in his mid-30s. By the conventional CV-reader's standard he's mid-career, no logo factory employers, no acquisition, no liquidity event. His LinkedIn has 297 followers and a short About section. The CV is thin. Maybe others would see his CV, quickly scan and see a mid-level engineer and pass over him. I might even do that, had our paths not crossed through our original mastermind.
He's also the best engineer I know. If there was a way to invest a portion of my own money on Jan having an exit in the next ten years, I'd put a lot of it down. I'm betting on what he hasn't done yet.
What the pair looks like
I'm not the one teaching Jan anything on the engineering side. I'm bringing him because on the engineering side he can teach me. I'm bringing the customer, the writing, the distribution, the twenty years of running small businesses, and the willingness to make twenty decisions a day that other people would still be researching. He's bringing the technical depth I do not have and the build cadence I cannot match alone. I was telling a prospect about bringing on a new CTO, and I said between him and me, I would be surprised if we couldn't handle 95% of the problems in the space that were thrown at us. Of course not all, but I would think most and I am not afraid to say what I can handle and what I can't.
That pairing is the post I'll write another time. What he brings, what I bring, why two people in this configuration ship more than ten contractors stacked together. For today, the headline is: I asked, he said yes, and the company has a CTO.
If you're about to hire your first engineer
Don't optimise for the CV. Look for the year of behaviour you've already watched. If you don't have a year of behaviour to look at, give a test task before you make the offer. Most founders run that order in reverse and pay for it.
The first 47 days of building in public were me and the models.
Monthly Revenues $11,800 | Clients 2 | Prospects 1 | Team: Me + Jan (CTO)
Day 48 of 365.