Learn Houston
The framework for AI-native workspaces.
An AI-native workspace is an app where the AI can take every action a human can take. The human can take every action the AI can take. Neither is a guest. Houston is a framework for building them. This guide explains what that means and how to build one.
Who this is for
Developers who have built apps with AI as a feature. You've added a "summarize with AI" button. You've called an LLM from a backend job. You've used an API.
This guide is about a different way of building. Not better for every case. But for the right case, it's a different product entirely.
What you'll learn
01 · 4 min read
The mental shift
Why AI as a feature has a ceiling, and what it means to build past
it.
02 · 3 min read
Humans and AI, as equals
The one architectural decision that makes everything else
possible.
03 · 3 min read
Pages and components
How a Houston app is organized, and how the AI reads it without
being told.
04 · 3 min read
Teaching, not configuring
How users bend a Houston app to their needs just by talking to it.
05 · 4 min read
Edge cases at runtime
How the AI handles cases you never wrote code for, without
waiting for you.
06 · 2 min read
Showing, not driving
How the AI points at things in the UI without hijacking the user.
07 · 4 min read
Building a Houston app
The shape of a Houston project, and how to ship one.
08 · 3 min read
What Houston isn't
The limits, the positioning, and when to pick a different tool.