What if planning your home didn't require expensive design software, specialised knowledge, or uploading every room photo to a third-party AI service?
That question inspired me to build ApnaGhar (เค เคชเคจเคพ เคเคฐ) โ a home-design application that brings room planning, coordinated interior styling, 2D layouts, 3D visualisation, and AI-assisted room understanding into one experience.
๐ Live Demo: https://apna-ghar.ai.studio/
๐ป Source Code: https://github.com/DEVDESAGAR1/apnaGhar
๐ค The Idea: Building for a Friend
The idea behind ApnaGhar started when my friend wanted to design their house. Choosing wall colours, coordinating flooring, arranging furniture, and imagining the final result can be challenging without the right tools.
I wanted to make that process easier by creating a space where my friend could explore interior styles, experiment with colours, and visualise room layouts before making real-world decisions.
That thought became ApnaGhar โ a project built around a simple idea: making home design more accessible through technology.
โจ What Is ApnaGhar?
ApnaGhar brings together three main experiences:
- Visualise: Explore rooms through interactive 2D planning and 3D visualisation.
- Coordinate: Experiment with architectural paint collections, colour combinations, interior styles, accent walls, trim, and flooring.
- Understand: Explore AI-assisted room-photo analysis and furniture recommendations.
The goal is to connect creative design with practical room planning.
๐ ๏ธ What I've Built
๐จ A Whole-Home Design System
Curated interior styles, architectural paint collections, colour-family tagging, coordinated wall and flooring finishes, and room-level and whole-home design controls.
๐ท AI-Assisted Room Understanding
Image inspection, indoor-room suitability validation, and furniture-review recommendations, with a tested local vision-processing workflow using Gemma and Ollama.
๐งฑ Interactive 2D and 3D Planning
A 2D room planner and Three.js-based 3D viewer that bring room layouts and material choices together.
๐ Privacy-Conscious by Design
ApnaGhar explores local AI processing to help users maintain control over personal room photographs and reduce unnecessary dependence on external AI services.
๐งฐ Technology Stack
- React, TypeScript, and Vite
- Three.js for 3D visualisation
- Interactive 2D room planning
- Gemma and Ollama for local AI-assisted room understanding
Development checks included successful TypeScript checks and a production build with no reported errors.
๐ฑ Why Open Innovation Matters
Open-weight models give developers the freedom to experiment, adapt technology, and build useful applications without depending entirely on closed AI services.
With ApnaGhar, I wanted to explore how open-source AI could support everyday home-design decisions while keeping privacy and practical reliability in mind.
๐ What's Next?
I want to continue improving room understanding, furniture recommendations, 2D planning, 3D visualisation, and the overall user experience based on real-world feedback.
โค๏ธ Built for a Friend, Open to Everyone
ApnaGhar started with my friend's need to design their house and grew into an opportunity to explore how software engineering and open-source AI could make home planning easier.
For me, this project is about using technology to solve an everyday problem and helping people feel more confident about the place they call home.
๐ Try ApnaGhar: https://apna-ghar.ai.studio/
๐ป Explore the code: https://github.com/DEVDESAGAR1/apnaGhar
Thank you for checking out my project! I'd love to hear your feedback and ideas for improvement.
๐ Challenge
Hacktoberfest DEV Weekend Challenge โ Build for a Friend
Category: Best Use of Gemma
Top comments (0)