Real estate professionals using VR Maker to create virtual tours faced a persistent problem: most properties are photographed while still occupied. Panoramic images captured cluttered rooms, outdated furniture, and personal belongings that distracted from the space itself. Professional physical staging costs thousands per property and is not scalable. Users needed a way to transform captured panoramas into clean, aspirational presentations without re-shooting.
The design challenge was making AI-powered image transformation feel trustworthy and controllable in a professional context. Real estate agents need to present properties accurately while highlighting potential — the line between enhancement and misrepresentation matters. The AI staging feature needed to offer style choices, allow regeneration when results were not right, preserve a history of generated versions, and integrate naturally into the existing VR Maker viewer experience rather than feeling like a bolted-on tool.
I mapped the real estate photography workflow to understand where AI staging fits naturally. The key insight: agents do not want to edit images — they want to present options. I designed the feature as a presentation tool rather than an editing tool: select a room, choose a style, generate, and compare. This framing made the AI feel like a styling assistant rather than a photo manipulation tool.
I designed a style picker that lets users choose desired aesthetics (modern, Scandinavian, classic, etc.) before generation. The AI produces multiple style variations from the same source image, and users can regenerate individual results without losing previous versions. A generation history preserves every version, so users can compare and select the best representation of the space.
Rather than isolating Deco AI as a separate tool, I integrated it directly into the VR Maker viewer experience. Viewers can access a dropdown to switch between the original panorama and multiple AI-styled versions, making the comparison immediate and contextual. This integration made the AI-generated content feel like a natural part of the tour rather than an afterthought.
Deco AI was designed as a premium add-on purchase with credit-based usage. I designed the purchase and credit management flow to be transparent — users see exactly how many generations they have remaining, get notified before running low, and can top up without leaving their workflow. The goal was to make the premium feel valuable, not restrictive.
Positioning Deco AI as a way to present style options rather than manipulate images changed how users perceived the feature. Real estate agents could show clients potential rather than fabricating reality — an important ethical distinction in property marketing. This framing also simplified the UX: no complex editing controls, just style selection and comparison.
Integrating AI results directly into the VR Maker viewer meant that the generated content lives alongside the original. Tour viewers can toggle between reality and possibility, which builds trust — the original is always accessible. A separate workflow would have disconnected the AI output from its context.
Every generation is preserved. Users never lose a version they liked, and regenerating does not overwrite previous results. This non-destructive approach reduces the anxiety of using AI tools — there is no wrong choice because nothing is permanent. The 70% utilization rate suggests users feel confident experimenting with their credits.
Deco AI launched as a premium add-on for VR Maker, creating a new revenue stream while solving a genuine pain point for real estate professionals. The feature integrated directly into the existing viewer experience, making AI-generated staging feel like a natural extension of the virtual tour rather than a separate product.
This project taught me that AI feature design is fundamentally about trust. Users do not need to understand how the AI works — they need to trust the results and feel in control of the output. The non-destructive history, transparent credit system, and presentation-over-editing framing all serve the same goal: making AI feel like a reliable professional tool. The high utilization rate tells me we got the trust equation right.