The Designer as Intelligence Layer

When Modsy and Havenly tried to democratize interior design, they made a fundamental mistake: they tried to replace the designer with an algorithm. Feed in room dimensions, select a style quiz result, get a 3D render with shoppable links. It was convenient. It was also generic.

The problem isn't that AI can't help with furniture selection — it's that the AI was trained on the wrong signal. Popularity metrics, purchase history, and visual similarity are useful, but they miss what makes a great furniture choice: context. A designer doesn't just see a chair. They see how walnut ages in southern light, why that particular joint method matters for longevity, how the proportions relate to the other pieces in the room.

Learning From Professionals

Patina's AI doesn't learn from what's popular. It learns from what professional designers actually select, and why. Every time a designer in our Founding 50 cohort classifies a piece — tagging its material honesty, construction quality, design lineage, spatial personality — that knowledge enters the system.

Over time, the platform develops what we call 'designer-grade taste' — the ability to recommend pieces not just by style, but by the deeper qualities that professionals evaluate instinctively. Material integrity. Proportion harmony. Aging character. Maker reputation.

No Moat Without Taste

Anyone can build a furniture marketplace. The technology is commodity. What isn't commodity is a classification system built by 50 working designers who stake their professional reputation on the quality of their selections. That's Patina's moat — and it's why we're starting with designers, not consumers.