Skip to content
The Post-Build Cliff

Where learning ends, hiring begins

The founder skill in an AI-prevalent world is knowing exactly where your competence should stop and someone else’s should start. This module is about spotting that line and walking across it without getting taken.

The premise

Most founder advice on hiring engineers is written for people who can hire engineers — i.e., people who know enough engineering to evaluate a candidate. That’s not you yet. You’re hiring help in a domain where you can’t out-code, out-architect, or out-debug the person you’re evaluating.

That sounds bad. It isn’t. The skills you do have — judgement, communication, business context, knowing what your users want — are exactly the skills that pick a great engineering hire over a great-looking-on-paper one. You don’t need to interview them on data structures. You need to interview them on judgement.

What this module covers

Three lessons, one decision framework each.

01. Five signals you’ve hit the line

Concrete observations from your own work that tell you it’s time. Not feelings. Five of them; if three apply, you’re past the line and didn’t notice.

02. Who to hire when

A matrix of contractor / fractional CTO / agency / full-time engineer — what each is good for, what each costs, where the traps are. With heuristics for a non-technical founder spending real money for the first time.

03. Four interview questions

The four questions that surface judgement, communication, and self-awareness in 30 minutes. Specifically designed to be useful when you can’t evaluate code.

The single thesis

You don’t need to become an engineer. You need to know exactly where your competence should stop. Walking across the line is a skill, not a defeat. The founders who do it well end up shipping more than the ones who try to do everything themselves.

Skills you’ll pick up in this module

  • Recognising your own “I’m at the line” signals before they cost you a weekend
  • Picking the right kind of help for the stage and shape of your problem
  • Conducting a 30-minute first interview that surfaces real signal without coding
  • Negotiating scope and rate without being either soft or paranoid

References & further reading