Show the path.
Make steps, sources, and uncertainty easier to see and question.
Exploring a more legible, more steerable way to work with AI.
What if an AI workspace made its structure visible—so people could understand the path from intention to action, and shape that path as they go?
Aurora is in progress. The following model is a design hypothesis, not a description of released software.Select a layer to see the question behind it. Every element shown here is proposed or conceptual.
A legible path from intention to action.
The proposed orchestration layer would organize steps, handoffs, and checkpoints so a person can understand what is happening and redirect it.
How can complex work remain steerable?
SELECT A LAYER · TAB AND ENTER WORK TOO · CONCEPT ONLY
Make steps, sources, and uncertainty easier to see and question.
Design for correction, choice, and appropriate human checkpoints.
Let prototypes and feedback change the model as the work develops.
Aurora is a direction under investigation. This page will change as its ideas become prototypes, lessons, or working features.
Back to projects