Tech stack

A tech stack is the set of technologies your product is built on: the language and framework for the front end, the back end and database, the hosting, and the services bolted on for auth, payments and the like. It is the foundation every future decision sits on, so choosing it well early saves you from expensive rewrites later.

For a lean founder building with AI, the smartest stack is a boring, popular one. Models are far better at writing React, TypeScript, Postgres and the common frameworks than at niche tools, simply because they have seen vastly more of that code. A mainstream stack means your AI agent makes fewer mistakes, more libraries solve your problems out of the box, and you can hire help when you need it.

It is easier to picture with real tools. Say you are building a SaaS app with Claude writing most of the code: because it has read millions of React and TypeScript projects, it produces working code far more reliably than it would for some obscure language. Say your database is Supabase, which is Postgres with auth and APIs already wired in, so you skip building the plumbing yourself. And say your marketing site sits on Webflow instead of custom code, so you can ship and edit pages without a developer.

Resist the urge to pick something clever or cutting-edge for its own sake. The right answer is usually the proven, well-documented default, because that is where both the ecosystem and the AI are strongest.