Bring Your Own AI
If the application owns the semantics, why does it also need to own the intelligence? Welcome agents as first-class users: the application becomes stable, the intelligence becomes portable.
Hi, I'm Luc Bocahut. I'm a systems architect: I design and build software meant to run for years, from trading infrastructure to personal agents. The ideas here are not commentary; they emerge from building products and running them in production. More about how I think.
If the application owns the semantics, why does it also need to own the intelligence? Welcome agents as first-class users: the application becomes stable, the intelligence becomes portable.
How we build products at Commons now. Ideas become issues, agents turn them into merged, deployed software, and the default state is shipping.
A longer companion to the manifesto on regenerative software.
Family logistics scatter across apps, inboxes, and memory; Anton is a personal agent OS that runs them in production from hardware at home.
Running Lightning nodes means reacting to events no one is watching for; a node manager UI with webhook-driven event and command handling.
A community grocer's shoppers lost track of their cart between categories; a continuous-scroll shop that keeps the cart in view without page reloads.
The Grand Paris Metro excavates earth at 60+ stations over 13 years, and every load must be traced; an isomorphic web app tracks it all.
Cavalry's operators needed one surface to run prospect pipelines, customer health, and outbound; Operate was that dashboard.
Every trade has to reconcile against the broker's version of the truth; Tiverton's reporting and reconciliation pipeline, still running 20+ years on.