3D AI web application (MVP)

I’m building a browser-based application that turns any natural-language prompt into a downloadable 3D asset. When a user types something like “realistic 3D globe with textures,” the back-end should call the Meshy API, return a .glb or .obj with textures, show the result in a WebGL preview, and let the user download the file. The interface needs to stay minimal—prompt field, progress indicator, real-time viewport, and download button—while remaining fully responsive on mobile and desktop. Account creation will use a straightforward email/password flow; OAuth providers can be added later, but email must work flawlessly from day one. I’ll deploy the MVP to my own cloud account (AWS is ideal, though any major provider works) and would like infrastructure-as-code plus a clear deployment script so I can repeat the set-up in other regions. Deliverables • Live MVP accessible via browser and mobile • Complete, well-commented source code • README covering local set-up, environment variables, and deployment steps • Short guide on maintenance and future scaling strategies • Quick hand-off session to walk through the code and infra Acceptance criteria: a prompt within Meshy’s free-tier limits generates a model, previews correctly in the viewer, downloads successfully, and opens in Blender without errors.

Project ID: 39737123