How to Vet a Founding Engineer Role Before You Bet Your Career on It
DexDue diligenceAnt McCabe, VP Commercial8 min read

One in five roles we run at Dex right now is for a founding engineer. I haven't seen a market with that skew before, and many of the engineers we talk to are increasingly asking for exactly that kind of ownership.
Others are going a step further and skipping the job altogether to build something of their own instead. And while the AI startup boom is what's driving all of it, accelerators like Y Combinator, HF0, Speedrun and Entrepreneur First have gotten so good at turning engineers into founders that people who'd have laughed at the idea two years ago are now taking the leap.
The upshot is more companies than ever need someone to build the first version and make the early calls on architecture. Barely any of these roles come with the regular scaffolding, HR team or formal interview rubric anyone bothered to write down. And when everyone’s doing it for the first time, there’s no one around to ask ‘hang on a second, is this normal?’
What you get instead is a process the founders have built themselves, often from scratch. How they run them tells you a lot about how they run everything else, including whether they’re organised, how they make decisions and how they treat people. That’s worth paying attention to, because a founding role is a sizable bet on a small group of people and the interviews are the best look you’ll get at them before you commit.
Reading the loop
No two founding engineer interview loops look alike. What shapes the process more than anything else is who's already on the founding team.
Non-technical founders can't assess your system-design chops themselves, so they might lean on someone who can, usually an advisor, an investor or a consultant they trust. Technical founders skip the middleman and run the process on their own instincts about what good looks like. The same interview format can mean two completely different things depending on who's sitting across from you.
Here's the thing I see candidates get backwards all the time. If a non-technical founder never asks you a single technical question, that's not a red flag. The real question is whether they've even worked out how they're assessing you at all, whether it's someone else in the room, a trial task reviewed by an advisor, anything. A founder who's thought about that has thought about how to hire well. A founder who hasn't is winging the most important decision they'll make this year. And you're the guinea pig.
That plan, or the absence of one, tends to show up in what they spend the conversation testing for. A founder who knows they can't judge your code will often deliberately test something else instead: communication style, gravitas, whether they'd put you in front of a VC and trust you not to embarrass them. That's fair, given their own fundability is what decides whether your job exists in twelve months.
The tell is whether they can say that's what they're doing and why. A founder with a plan has an answer ready. A founder without one changes the subject, or falls back on 'we just wanted to get to know you', which is another way of saying they never worked it out.
Many founders lean hard on the pitch of autonomy, impact and ownership because it's exactly what this candidate pool wants to hear. It’s not misplaced either. Lack of autonomy is the biggest reason candidates in Dex's own data walk away from an offer, well ahead of anything else.
Watch whether their conviction survives you pushing on it. Every founder genuinely believes they're building the next big thing. That's the whole basis of the role. A founder who folds the moment you push on the vision or what you'd really own is telling you both were thinner than they sounded.
You don’t need to play devil’s advocate to find out. Asking what could kill the company in the next year, or what they’ve changed their mind on since starting, gets you there without putting anyone on the defensive. Good founders tend to enjoy those questions and the ones who dodge them with a pitch-deck answer are telling you something too.
Speed is also a key signal, in both directions. The founding-engineer placements that close fastest move in two to three weeks. Much slower than that and you're right to wonder whether the company's being careful, or just doesn't have anyone driving the process.
But the plan question comes first. A founder with a great pitch and real conviction, but no idea how to assess you, is still winging your hire. Read for that before anything else.
What to ask the company
I remember when asking a founder about runway felt like asking someone their salary at a dinner party. Those days are over. Serious founding candidates have every right to find out about cash reserves, current burn, customers signed, design partners in the pipeline or whatever tells you this thing has legs beyond the deck.
In fact, not asking is the thing that looks unprepared now. So ask directly and plainly, like you're checking a spec sheet rather than trying to draw out a confession.
- Runway. The round size on its own tells you little. VC-backed London startups typically raise £1–5m at pre-seed and seed, and the strongest AI companies raise well above that. What matters is whether the runway outlasts the gap to the next raise, and that gap is growing. UK companies took a median of 14.4 months to raise again after seed in 2025, up from 12.4. Ask what hiring plan the runway assumes and when they plan to raise next. Too little runway is the obvious worry, but too long (say 18 months) can mean a company isn’t spending fast enough to grow.
- Technical assessment. If someone other than the founder assessed you technically, ask what they told the founder afterward. Same test as section one, where a specific answer means there was a real plan behind it. That could be a written note, a clear yes or no or a named concern about your work. If they said ‘you seemed strong’ that usually means there wasn't much of an assessment behind it at all.
- Product-market fit. Ask the founder whether they think they’ve found product-market fit yet and what they’re still figuring out. At this stage, no startup has everything solved and the best founders know that. They’ll tell you what’s working, what isn’t and what they’re testing next. If a founder is adamant they’ve already cracked it, that’s worth a follow-up question or two. It might mean they’re reading too much into early signals, which can make it harder to adjust when things change.
- Equity. Our latest report, based on 600+ conversations with engineers over a three-month period, found senior candidates are targeting £100,000 to £150,000+ base salary with equity on top – though plenty say they’d be willing to take a big cut for meaningful ownership. One candidate was ready to give up two-thirds of their salary for the right equity stake. If you’re making that trade, look closely at the vesting. Most grants vest over four years with a one-year cliff, so ask what happens if things change before then and get the answer in writing before you sign.
It’s normal for founders to get a little protective when they’re pitching their own company, so what you’re looking for is how they handle a direct question. The founders worth betting on can talk through market size, runway and traction without flinching, and they’re just as comfortable telling you what they don’t know yet as what they do. You’ll often find they’ve thought about the hard questions more than you have and enjoy the chance to dig into them. That kind of clear thinking under pressure is a good sign the vision has real substance behind it.
The current strength of the market means you have more leverage in that room than you think. Founders know you're probably deciding between this opportunity and several others. You could even be considering creating the thing yourself.
Any founder serious about getting the top 10% of engineering talent already knows that. If they act like you're being difficult for asking, they're telling you they haven't had to compete for good people in a while, or don't think they should have to.
What to find out on your own
Most of what separates a candidate who's done the prep from one who hasn't isn't exactly glamorous.
Simple things make a difference, like knowing exactly who you're meeting before you walk in and already having an understanding of what they've built. Who's backing them and what they've said publicly about where the project is going. None of that's hard to find, but it is easy to skip.
I recently had a candidate going for a founding CTO role who ran reference checks on the two founders she was interviewing with. If a founder's raised money before, run a team before or sold something before, there are people out there who watched them do it up close. A former cofounder, an investor from an earlier round or someone who worked for them before this company. Would they work for this person again? Find them and ask.
Even after six or seven hours of contact in the interview process, you still haven't spent a real week with these people. You haven't seen what happens when a demo breaks the night before a customer call, or the founder's exhausted and short with everyone at 11pm because the last sprint went sideways. At some point you're taking a leap of faith and no amount of searching changes that.
But that's no reason to skip the research.
Taking the leap
Put all three of these steps together and you get about as much certainty as a decision like this ever offers. Due diligence will never get you all the way there, and that's fine.
A founder saying yes to you is betting on someone they've known for a matter of weeks, going on how you handled a few hard questions and what a couple of strangers said about you secondhand. Everyone in this process is taking some version of the same leap.
The only real mistake is skipping the parts you can control. Read whether they had a plan for you. Find out what you're entitled to know. Check what's checkable.
There's no version of this where you know everything for certain going in. There's only a version where you prepared properly, and a version where you didn't.


