Ambient Interfaces
What a product looks like when the interface assembles itself around the task instead of the menu.
Projects
Apps, agents, experiments, client builds and internal products. Where we can, we publish the decisions behind them — the useful part is rarely the screenshot.
Some of this work is commercial and some of it is curiosity. We keep both visible because they inform each other constantly.
Client work teaches us what breaks under real constraints. Experiments teach us what's possible before a client asks for it. The Academy teaches whatever survived both.
This archive is built to grow into a searchable body of work: write-ups, tutorials, teardowns and post-mortems.
What this looks like
Products we own and operate, listed in full on the Apps page.
Agent builds against specific workflows with measurable outcomes.
Labs work: interfaces, protocols, benchmarks and failures.
Selected Studio engagements, shared with permission.
Tooling we built for ourselves that turned out to matter.
The reasoning, the trade-offs and the parts that didn't work.
A growing archive. New entries are added as builds ship and experiments conclude.
What a product looks like when the interface assembles itself around the task instead of the menu.
A working pattern for when an agent should stop, summarize and give control back to a human.
A public log of ideas taken from prompt to deployed product inside five working days.
Client builds where AI wraps around software that is twenty years old and not going anywhere.
Our internal scoring rig for deciding whether a model change actually improved the product.
Teaching materials built from real builds, not toy examples. Feeds directly into the Academy.
Studio takes on a small number of builds at a time. Tell us what you're trying to make exist.