How the work got here
I started by gluing websites together with plugins. The useful skill was knowing what did each job well and not rebuilding capability that already existed.
The tools changed. Plugins became APIs, cloud services, infrastructure, models, skills and agents. Platform engineering became a way to compose work. One coding agent became several.
Then the agents worked across more than one product at a time. I needed a system to understand what had changed, what had survived review and what the software had actually proved.
Woodhouse is that public record. The projects are real. The claims stay smaller than the evidence. The failures explain which factory controls came next.
What I build
Systems that move an idea toward a running product: applications, infrastructure, release paths, agent workflows and the controls that keep their claims honest.
The projects are varied on purpose. Games expose product and visual judgement. Media systems expose catalog, playback and delivery. Reconstruction work tests whether an agent can stay inside source evidence. MAX tests where software proof stops and physical proof begins.
Read the full founding account, I Didn't Mean to Build a Software Factory, for the system behind the work.
The work should carry the résumé
Woodhouse is a public record of the products, their architecture and the factory lessons that survive. It is not a list of unsupported adjectives. Start with the projects, then follow the incidents and architecture.