Inspiration
As homeowners, we struggled to visualize our space’s potential—hiring designers was costly, and DIY tools lacked realism. We wanted to make professional interior design accessible to everyone, so we built RoomDesign AI to bridge that gap.
What it does
RoomDesign AI turns empty room photos into detailed 3D designs. Users select styles (Modern, Nordic, etc.), customize furniture/colors in real time, get a shoppable list of items, and connect with local renovation contractors. Free tier offers basic access; Premium unlocks full features.
How we built it
We combined JavaScript/React for the frontend, Python/TensorFlow for AI room analysis, and Three.js for 3D rendering. AWS hosted the backend, while Google Maps API powered contractor location matching. We iterated on UI/UX to ensure real-time customization felt seamless.
Challenges we ran into
AI struggled with low-light/obscured room photos—we refined image analysis models to prioritize layout/dimension detection. Real-time 3D preview lagged on lower-end devices, so we optimized asset loading for faster performance. Contractor data matching also required balancing location accuracy with result relevance.
Accomplishments that we're proud of
We created a tool that generates 3D designs in minutes (vs. hours/days with traditional tools). The real-time customization feature works flawlessly, and users love the shoppable list that simplifies turning designs into reality.
What we learned
We mastered balancing AI complexity with user simplicity—even advanced tech needs an intuitive interface. We also learned local market nuances matter: contractor matching needed region-specific budget/pricing logic for accuracy.
What's next for this project
We’ll add mobile responsiveness, integrate more design styles, and partner with furniture retailers for live inventory links. We also plan to build a user account system to save designs and expand contractor matching to more regions.
Log in or sign up for Devpost to join the conversation.