July 2026
The same living room shown as Modern Indian, Japandi and Garden House, with the broader experiment around it.
I could tell when I liked a sofa in a shop. I had almost no idea whether I would like it next to a particular rug, under a particular light, in a room I had only seen empty.
When I started looking at interiors for our new home, I had zero intuition for how one colour would go with another. I did not know what shape or size of rug would work with a sofa, whether individually nice pieces would look coherent together, or how their proportions would translate into my actual space.
Visiting more shops gave me more things to like. It did not necessarily make the room easier to imagine.
INTERIOR CONCEPT STUDY · MELBOURNE
6 rooms×4 directions×4 looks= 96 studies
Waurnvale Interiors started as a way to see possibilities in the house rather than in isolation. I used photographs of the actual rooms, the floorplan, pieces I liked and a growing collection of references to build an interactive design study.
The result explored four broader design worlds—Modern Indian, Mid-century, Japandi and Garden House—across six rooms. Each room had four increasingly different treatments, which created 96 selectable looks rather quickly.
I could move between the existing room, a selected concept and the wider family of options. The experiment also kept track of palettes, furniture ideas, lighting, planting, joinery and the measurements that still needed to be confirmed.
AI gave me cheap visual iteration. Instead of trying to imagine whether a mustard wall, walnut bed, striped sofa or chequered threshold might work, I could place the idea into something resembling the real room and react to it.
That reaction was valuable even when the image was not perfectly accurate. I learnt what I repeatedly gravitated towards, which combinations felt too coordinated, and where a single interesting piece needed the rest of the room to become quieter.
The references were translated rather than copied. A saved image could contribute a material, a lighting idea or a mood without becoming a request to reproduce someone else’s room.
The project gave me a vocabulary for interiors before it gave me a shopping list. I could discuss proportions, palettes and the role of a room with much more clarity—and rule out quite a few expensive ideas before buying them.
The visuals are still concepts, not construction drawings. Dimensions, colours, products and concealed building conditions need to be checked in person. AI helped me fall in love with a direction; measurements and trades still have to de-risk it.