Learning by making.

An agent-assisted reading desk that turns a noisy set of sources into a traceable daily brief.

Exploring how citation-first interfaces change trust in generated synthesis.

A structured conversation that helps teams turn an early idea into a sharper product brief.

Testing whether a model can surface assumptions without making the process feel like a form.

A plain-language comparison tool for choosing a model based on the job, not the leaderboard.

Looking for better ways to communicate tradeoffs without hiding meaningful technical detail.