Oculan.ai started with a very simple frustration: AI image generators are remarkably good at making pictures, but I kept wanting the thing after the picture—the actual graphic. If I generated something for a presentation or document, I didn’t want a flattened rectangle of pixels. I wanted to move a shape, change a color, tweak a gradient, delete an element, or just drop the finished artwork straight into PowerPoint and get on with my day. What began as “surely AI can generate editable graphics” turned into a much deeper rabbit hole involving vectorization, semantic structure, layering, geometry, and an unreasonable number of experiments. The approach evolved from trying to make image generation behave like a drawing tool into building a system specifically around structured, editable artwork. The goal throughout has stayed pretty simple: generate something polished enough that you probably won’t need to edit it—but make sure you can. That became Oculan.ai.

Oculan.ai

Scalable, fully editable, document-ready graphics

Scalable, fully editable, document-ready graphics