Turn Your White Box into a Cozy Space in Seconds.

Uzbekistan's first AI designer that uses real furniture from local stores (Dafna, AIKO). Visualize, budget, and buy.

View Demo Now

Before: Empty White Box

After: AI Design

The Problem We are Solving

Moving into a new apartment in Uzbekistan shouldn't feel like starting from zero.

The "White Box" Trap

The Dream vs. The Concrete Reality

New high-rise apartments are delivered as empty concrete shells, leaving 100,000+ homeowners to start renovation from zero.

The Visualization Gap

Pinterest to the Bazaar

Inspiration (Pinterest/Instagram) clashes with reality. You can’t find the right furniture locally, leading to the "Cost of Guessing".

The Cost of Guessing

Millions Wasted

Professional designers charge $1,500+ for drawings. Without a plan, homeowners spend millions of Soums on errors (too big, clashing colors, returns).

How It Works

Three simple steps to transform your space from empty concrete to a beautiful home.

Upload Photo

Take a photo of your empty room and upload it to our platform.

Choose Style & Budget

Select from Modern, National, or Hi-Tech styles and set your budget.

Get Your Design

AI generates your look with a complete shopping list in UZS.

Under the Hood

Built with cutting-edge technology for real-time AI interior design.

Multimodal RAG

Retrieval-Augmented Generation for intelligent furniture matching (using image, text, and room dimensions).

Vector Search

Semantic search across thousands of real furniture items, ensuring highly relevant, budget-aligned recommendations.

Microservices

Scalable architecture that separates the AI engine (Python) from the core business logic (Java), allowing independent scaling and robust deployment.

Core Tech Stack

Java Spring BootCore Backend

Used for building a robust and scalable microservice architecture, managing business logic, user authentication, and secure data handling.

Python FastAPIAI Engine

Selected for its high performance and asynchronous capabilities, providing a fast and efficient endpoint to expose the Multimodal RAG and Vector Search models.

Typescript/React/Next.jsFrontend

Leveraged for modern, type-safe development to create a fast, server-rendered, and highly interactive user experience (UI).

Our Roadmap

From hackathon prototype to the go-to platform for home renovation in Uzbekistan.

Current Prototype & Image Output

Current

Core AI functionality: Users can upload an image of a room and receive a redesigned 2D image output. This phase validates the AI selection logic and furniture catalog integration.

November-December 2025

MVP & Interactive 3D Model Integration

The shift to MVP will include the interactive, interactive 3D isometric room model output, filled with furniture items from major local retailers (Dafna, AIKO, and more).

December 2025