How we hire at Terac

We're a small team building something ambitious, and we hire deliberately. The next twenty hires will set the tone for the next two hundred. So we don't hire to fill seats. We hire when we find someone we'd genuinely fight to keep.

This post is the un-mysterious version of our process. What we actually pay attention to, who you'll talk to, and how the work trial works. If you're considering applying, or you've already started talking to us, this should answer most of the questions you'd otherwise have to ask.

How we think about hiring

Speed is a tradeoff we're willing to make. A bad hire on a ten-person team isn't a small problem to absorb later. It changes how the next person gets evaluated, what conversations happen in Slack, and how much trust people extend by default. So we move slowly, and the bar for "yes" sits well above "this person is fine."

We pay much more attention to trajectory than to pedigree. A senior title from a name-brand company tells us very little about whether someone will be excellent here. What does tell us something: how fast you've grown in your last role, what you taught yourself without being asked, what you built when no one was watching.

A note on interviews. They're a flawed instrument. Some people are excellent under interview pressure and underwhelming on real work; others are the opposite. The work trial exists to correct for this. A few hours of conversation, no matter how sharp, can't tell you what someone is actually like to build with. A week of shipping together can.

What we're looking for

The specifics shift by role, but a few things show up in almost everyone we end up hiring:

A founder somewhere in your past. Not a requirement, but it's our favorite signal. Anyone who's tried to start something (a company that didn't work, a side project that found a few users, a community you ran in college) has been forced to develop instincts that are hard to fake. You've sat with ambiguity. You've shipped something embarrassing because shipping something was the point. You know what it costs to actually own an outcome.

Agency that doesn't need permission. The default mode here is: notice the thing, decide it matters, go fix it. Not file a ticket, not wait for a planning meeting, not ask three people first. We trust people to make calls and accept that some of those calls will be wrong. What we don't tolerate is sitting still while a problem gets worse because no one was officially assigned to it.

Genuine interest in the problem. We can teach a lot of things, but we can't teach someone to care. The people who thrive here have already been thinking about this space. About how human work and AI work fit together, about what frontier labs actually need from a labor layer, about why panel quality is so hard. If you've read the team's writing and have opinions about where we're wrong, that's a green flag.

Real depth in something. Whatever your craft is (code, design, distribution, ops) we want to talk to people who've gone deep enough to have strong, sometimes unfashionable opinions. Generalists are great; generalists who are also excellent at one specific thing are who we want to hire.

Always tinkering on something. Not at your day job. The people we love hiring tend to have a half-finished side project open in another tab. A weird little tool they built for themselves. A repo they push to on Sundays. A Figma file no one asked for. It doesn't have to be impressive or even working. The signal we care about is that you're curious enough to build things in your free time, because that almost always translates into how you show up at work.

The process

Most candidates go through roughly this sequence. Not every role is identical, but the shape is consistent.

Application or intro. Either you apply through the site, someone refers you, or we reach out. From there it's the same path either way. Every application gets read by a human.

A call with Zac (Co-founder, CEO). First conversation. Zac wants to understand your background and what you're optimizing for in your next role; you should be asking him hard questions about the company.

A call with Jack (Co-founder, CTO). A separate, role-specific conversation. For engineers and PMs, this usually means walking through real problems together. For everyone else, it's a deeper look at the parts of the product and tech surface you'd be working alongside.

A call with an engineer on the team. This one isn't with a founder, on purpose. It's the call where you find out what the day-to-day is actually like. The rhythm, the tools, the things that are great, the things that are still rough. Ask the questions you wouldn't ask us.

A paid, week-long work trial. Five working days on a real project, with a real slice of the team, shipping a real outcome. We pay full daily rates. We're flexible on scheduling; most candidates do this evenings/weekends or take time off from a current role. A week is long enough that you stop performing and we stop interviewing, and we both get to see what working together actually feels like. You'll get a written brief in advance with what we're looking for and what success looks like.

Debrief and decision. Once the trial wraps, everyone who worked with you writes up their take independently before we discuss as a group. We'll tell you that same day whether it's a fit or not.

What we owe you

A short, predictable process. Three conversations and a work trial. That's the standard shape, and we don't bolt on extra rounds because someone wants more signal.

Real money for real work. The trial is paid in full, in advance of the decision, and it's paid whether or not we hire you.

Direct feedback. If we pass after the interviews, we'll tell you why, even when it's awkward. We've been on the receiving end of vague rejections and don't want to do that to anyone else.

An honest read on the company. When we make an offer, we tell you the actual numbers: revenue, runway, growth, what your equity is plausibly worth in the realistic scenarios, not just the good one. Most companies are evasive here. We'd rather you sign up with eyes open.

On fit

Terac isn't for everyone, and that's fine. The people who do their best work here want a wide surface area, real ownership, and very little process. The people who struggle want clear weekly check-ins, a manager who tells them what good looks like, and a tight feedback loop they didn't have to build themselves.

Neither version is wrong. But the second version isn't going to be happy here, and we'd rather both find that out before the offer than after.

If the idea of being trusted to figure it out, and being held to a high bar for actually figuring it out, sounds like the kind of job you want, we'd love to talk. Our open roles live here.