According to Techstars, the one part of a startup you cannot bootstrap is deep human expertise. In this case, that is co-founder Dayna's 25 years as a college counselor.
It also includes her prior years as a therapist to high-schoolers, which no AI model or competitor can copy. Bootstrapping means building without outside money or hires. Techstars uses this story to show founders what still needs years of lived work when code itself has become cheap.
Table of Contents
- Who made the claim?
- What did he bootstrap in nine weeks?
- Why is expertise the moat?
- How do you turn expertise into product?
Who made the claim?
The claim comes from four-time founder Michael Taus. His companies include Rent.com and CrowdStreet among four startups, plus growth leadership at two more. He also mentors at Techstars, Founder Institute and StartOut, and states he is not an engineer.
That background shapes his point. He is writing as a non-technical builder who learned to ship with current AI tools. Techstars published his account on Oct 8 2026 in the Techstars founder essay.
What did he bootstrap in nine weeks?
Taus built Aiko, a live AI college advisor, in nine weeks. He worked part-time while helping build two other startups. He used no outside money and made no engineering hire.
The result runs in production as both an API and an app. It covers 2,561 American four-year universities with admissions math on top. Techstars describes the build in its account of the nine-week project.
Why is expertise the moat?
Everyone now has the same models, and building is nearly free. So the scarce asset is having something true to build against. A founder with real method starts ahead of a founder with only a model.
AI did not speed up thinking here, only building. The company reflects years of thinking against nine weeks of building. The question is what code is written against, not whether AI can write code, according to Techstars in its explanation of the moat.
How do you turn expertise into product?
The product inputs were raw transcribed counselor interviews, Dayna's method broken into repeatable moves, a versioned written source of truth, nine AI reviewers, and free federal college data. Four of the five cost mainly organization rather than cash. For other bootstrappers, the method is practical: When file and model disagree, the file wins.
- keep untidied transcripts so detail is preserved
- version the system-prompt voice as Dayna's real method
- make the written file beat the model when they disagree