Zhihuang Lin

His portfolio describes moving from frontend development into AI product management. It lists earlier work with API7.ai, SuperMagic and OpenCreator and current work on MuseOn, alongside independently published tools. That combination puts product decisions and the frontend mechanisms used to express them within the same body of work.[2]

Prepare the motion before the interaction

Oil Motion confirms key images, generates the transitions between them, checks the frames and prepares compressed browser assets. Scrolling, pointer movement, dragging or device orientation can control the resulting timeline. The browser plays prepared frames rather than asking a model to generate a new image for each movement.[3]

Choose an asset format for the actual display

The workflow distinguishes a rendered scene from reusable transparent material, and considers display size, random access and device memory when choosing a video or image atlas. It calls for a small pilot in the actual page before bulk generation. Frame problems such as changed anatomy or product structure require new source material; compression cannot repair an incorrect action.[3]

Oil UI begins with a product, audience and references, then makes explicit choices about typography, spacing and color. When several directions are requested, it asks for differences in composition rather than superficial recoloring. Its process uses desktop and mobile screenshots to inspect the result and asks the user to choose a direction before further refinement.[4]