Skip to content

ay-rod/encodeai-floorplanner

Repository files navigation

Kondo AI

Project Summary

Welcome to Kondo AI, the easiest and most entertaining way to design your home.

Kondo AI is a floorplanning app that seamlessly transforms real products to scale. It leverages GPT Vision to create an AI companion for personalised reviews and ratings, and gamifies the overall user experience. With accurate scaling of room sizes and furniture it also reduces the hassle of measurement-taking. We further simplify the planning process by tracking both the total costs and physical placement of wishlist furniture items in one place.

Running Locally

1. Clone your newly created repo:

git clone {{your-repo-name}}

2. Enter your newly created repo's directory:

cd {{your-repo-name}}

3. Install dependencies:

For yarn:

yarn install

4. Start the development server:

For yarn:

yarn run dev

5. Visit http://localhost:3000 in your browser to see the running app.


Project Overview

Problem Statement

The current home design process is cumbersome and stressful, relying on outdated methods like tape measures and imagination. This leads to potential costly mistakes in furniture selection. Consumers seek a reliable and efficient solution for home planning but are deterred by the tedious and time-consuming nature of existing floor planning software. We are here to provide a solution.

Our Solution Implementation

Kondo AI is a Next.js application, rendered on the frontend using HTML, Typescript, Javascript, CSS and IMGs. The backend is powered by Supabase.

We have designed web scraping algorithms that reliably find key product content (photos, price, colour, dimensions). LLMs then enhance and structure the scraped content into a JSON object, which becomes translated into an precisely scaled and moveable furniture item within the user's floor plan. These form building blocks the user can utilise to plan their layouts accurately, with real, to-scale products. Screenshots of their floor plans can be sent to GPT for item suggestions, layout review and budgeting advice.

Demo Images

Demo4 kondoAI Demo1

Releases

No releases published

Packages

No packages published

Contributors 5