This article was originally published on Steven’s Substack. Subscribe to his newsletter for more insights.

I’ve watched non-technical founders make the same hiring mistake over and over: they hire an engineer who interviews well, has an impressive resume — and months later, there’s no shipped product. Or worse, a product ships that customers find no value in.

So what went wrong?

  • Hired for technical depth, not shipping speed
  • Prioritized pedigree over product taste
  • Looked for “senior” when they needed “scrappy”
  • Didn’t pressure-test real-world problem-solving
  • Ignored communication skills entirely

This isn’t about “bad engineers.” It’s about hiring for the wrong things. And with AI changing how software gets built, the gap between “impressive on paper” and “actually useful for your startup” has never been wider.

This article is for founders about to make one of their most important early hires. Here’s what actually matters.

Iteration Speed Is Everything

Nothing matters more in a startup engineer than how fast they can ship something real.

SPEED REVEALS TRUTH

AI has compressed what used to take days into minutes. A competent engineer with the right tools can now prototype, test, and deploy faster than most teams can finish a design review.

This changes everything.

When feedback cycles shrink from weeks to hours, something magical happens: stakeholders actually see and feel the vision. And when they do, insights emerge that no amount of planning could have surfaced. Your brain works differently when you’re reacting to something real versus imagining something abstract.

If your engineer still needs multi-day review cycles to ship anything — that’s a red flag. In 2025, that pace isn’t just slow. It’s unacceptable.

SPEED COMPENSATES FOR ALMOST EVERYTHING ELSE

Miscommunication? Ship the wrong thing? No problem — take 10 more minutes and fix it. This is the hidden superpower of iteration speed: it becomes a currency you can trade for other mistakes. Misunderstood the requirement? Iterate. Picked the wrong approach? Iterate. Customer changed their mind? Iterate. Technical debt piling up? Rewrite it in a weekend. With AI, most products can be refactored or rebuilt in days — not months. The fear of “getting it wrong” disappears when fixing it is cheap.

But the reverse isn’t true. You cannot trade “great architecture” or “clean code” for speed. An engineer who writes beautiful code but takes three weeks to ship anything will kill your startup before it starts.

So, What Should You Actually Look For?

Knowing what matters is one thing. Actually filtering for it is another. Here’s what we look for at Worca — and what you should too.

60 HOURS IS THE BASELINE

This isn’t about glorifying hustle culture. It’s about math.

If iteration speed is your competitive advantage, you need engineers who can sustain high output. At early stage, there’s no slack in the system — every hour compounds. An engineer working 60 hours ships 50% more than one working 40. Over a quarter, that gap becomes insurmountable.

The best engineers don’t clock in and out. They’re obsessed with the problem. If someone’s optimizing for work-life balance before there’s a product, they’re not your person.

REAL-TIME COMMUNICATION IS NON-NEGOTIABLE

Co-location is ideal. Same timezone is the minimum — 100% working hour overlap.

Why? Because iteration speed dies in async. When you can’t get an answer for 12 hours, that’s 12 hours of blocked progress. Multiply that across a week, and you’ve lost days.

If you must go remote, the founder needs to over-communicate. Not “enough” communication — over-communication. Every assumption spelled out. Every decision documented. Every piece of feedback given immediately, not batched. In remote work, over-communication is just communication.

NO MANUAL CODING. PERIOD.

This is non-negotiable. Your engineer must build with AI.

Human fingers cannot move fast enough to compete with LLMs. This isn’t opinion — it’s physics. An engineer who manually writes every line of code is bringing a knife to a gunfight. They might write “cleaner” code, but they’ll ship 5x slower.

The engineers we look for treat AI as a multiplier, not a crutch. They prompt, review, iterate, and ship — in hours, not days.

SKIP THE PRODUCT DESIGN CYCLE

No PRDs. No product designers. Your engineer must be all-in-one.

Why? Because every handoff is a context transfer — and context transfer is where speed goes to die. When a PM writes a spec, hands it to a designer, who hands it to an engineer, you’ve already lost days. And worse, you’ve lost fidelity. The vision degrades with each translation.

The best early-stage engineers don’t need a spec handed to them. They talk to users, understand the problem, design the solution, and build it — all in one brain. Minimal context transfer, maximum speed.

If an engineer isn’t comfortable owning the full loop, they’re not ready for early stage. Skip.

RETHINK YOUR DEPLOYMENT PIPELINE

Traditional CI/CD is built for teams that ship weekly. You need to ship hourly.

Most pipelines are over-engineered — extensive test suites, staging environments, manual approvals, canary deployments. All designed to prevent mistakes. But at early stage, the cost of slow deployment far exceeds the cost of a bug.

The goal is minimum viable pipeline: essential tests only, straight to production. No staging environment you have to babysit. No approval chains. No waiting.

Every team is different — there’s no one-size-fits-all answer here. But if your deployment takes more than a few minutes, something is wrong. Strip it down. Remove every step that doesn’t directly prevent catastrophic failure. You can add guardrails later when you have something worth guarding.

The Bottom Line

Hiring your first engineer is one of the highest-leverage decisions you’ll make as a founder. Get it right, and you’ll have a partner who can turn your vision into reality — fast. Get it wrong, and you’ll lose months, burn cash, and end up back at square one.

The old playbook doesn’t work anymore. AI has changed what’s possible, and the engineers who thrive in this new world look different from the ones who succeeded five years ago. They’re fast, not precious. They ship, not polish. They own the full loop, not just their lane.

Find someone who moves like the company depends on it — because it does.

Ready to hire? We’ve created a job description template based on these principles. Copy it, adapt it, use it.

Want more thoughts on engineering, AI, and building teams? Subscribe to Steven’s Substack for regular insights.