Most people don’t lack taste. They lack tools.

We all have a vision for how our space should feel, but there’s a huge gap between imagining a room and confidently designing it. Sometimes, Pinterest is overwhelming, furniture is expensive, and you can’t actually see if your ideas will work together.

CURAiTE bridges that gap by acting as your personal interior design team, helping you visualize, refine, and bring your vision to life; curating the perfect space for any vibe.

• Gives smart room suggestions around cleanliness, organization, colour harmony, lighting, spacing and feng shui (the practice of arranging pieces in a living or workspace to create balance with the natural world)
• Lets you chat with an AI interior designer to fine-tune your space and ideas
• Recommends furniture pieces based off the AI recommendations you can actually buy
• Talk to a feng shui master to understand your room’s harmony
• Iterates on previous designs so each version builds on the last

Gemini API for image generation + interior design agents ElevenLabs for live voice agents with designers and feng shui experts SERP API for finding furniture online Frontend: Next.js, Tailwind CSS, shadcn Backend: Flask, SQLite

• Designing a UI that’s intuitive while still allowing deep customization

• Building a full-stack app end-to-end under time pressure
• Successfully integrating multiple AI APIs
• Turning abstract “vibes” into something visual and actionable

• The basics of feng shui :p
• How to integrate and orchestrate multiple APIs
• Prompt engineering for structured, reliable outputs

• Budget and location-aware furniture recommendations
• Deployment and stronger safeguards against malicious inputs

The app has a Next.js frontend (localhost:3000) and Flask backend (localhost:5001).

cd backend
python -m venv .venv
# Windows PowerShell
.venv\Scripts\Activate.ps1
pip install -r requirements.txt
copy .env.example .env

Then set your key in backend/.env:

GEMINI_API_KEY=your_key_here

cd backend
python app.py

Health check:

curl http://localhost:5001/api/health

From repo root:

npm install
npm run dev

Open http://localhost:3000.

Optional frontend env override (repo root .env.local):

NEXT_PUBLIC_API_BASE=http://localhost:5001

• POST /api/sessions (upload image)
• POST /api/sessions/:session_id/rate
• POST /api/sessions/:session_id/generate
• GET /api/jobs/:job_id
• GET /api/sessions/:session_id