I design computational systems for products that have to fit real bodies. Mass manufacturing pushes those products toward a single averaged shape: tooling is paid per geometry, so whichever population a mould or a last happens to fit becomes the standard. My work removes that constraint. Parametric generation removes the cost of creating variants, additive manufacturing removes the cost of tooling them, and one averaged standard becomes a few engineered ones, derived from how people actually vary.At REFRAMD, the eyewear company I founded, that took the form of a fit system built from thousands of 3D facial scans: three engineered bridge profiles instead of one, driven by a computational engine that turns a face scan into production-ready geometry in under seven seconds, scaled across more than forty products.The other half of the work is making these systems usable by people who aren't computational designers themselves. I teach Grasshopper through applied projects with design students and practising opticians, and I build the configurators that let non-technical colleagues drive complex geometry without opening the underlying definition. A system only creates leverage once someone else can operate it.