Voe AI is a browser-based AI video generation workspace for turning prompts, reference images, and frame controls into short-form video assets. The product is built around Veo 3.1-style workflows such as text-to-video, image-to-video, multi-image reference generation, first-and-last-frame control, generated audio, extendable shots, and natural-language scene direction.
The frontend is built with Next.js App Router and a custom marketing/homepage experience. The homepage includes a full-screen visual hero, prompt capsule, what-is-new section, core feature cards, outcomes grid, use-case explanation, a three-step creation flow, comparison table, pricing entry points, FAQ, and final CTA. Those sections are exactly the kind of structured content that can be managed cleanly in Sanity instead of being treated as one static page.
I am using Sanity as the content operations layer for the Voe AI project presentation and the educational material around the video workflow. Sanity is a good fit for managing homepage sections, feature copy, prompt examples, comparison-table notes, FAQ content, localized copy, demo metadata, and future tutorial pages that explain how creators can plan scenes, upload references, and iterate on video outputs.
The product itself separates content management from generation logic. The app handles authentication, credits, prompt input, reference-image and keyframe uploads, aspect-ratio choices, model options, generation state, polling, preview, and download. Sanity can own the surrounding product narrative: what each generation mode does, when to use it, which examples are featured, and how the workflow should be explained to users across different pages and locales.
The main challenge is presenting a powerful AI video tool without making the interface feel like professional editing software. Sanity helps keep that experience adaptable: I can update feature descriptions, launch notes, comparison content, FAQs, and tutorial blocks from Studio while keeping the Next.js code focused on the video-generation workspace. My next step is to expand the Sanity schema with demo video objects, prompt libraries, Studio screenshots, and localized guide content that can feed the public site directly from Content Lake.
