Decorly

Ever bought a couch you loved online, only to hate it in your living room? That feeling of uncertainty and potential regret is exactly what DECORLY eliminates. Our platform is your creative co-pilot.

Comment

Inspiration

The inspiration for DECORLY came from a universal frustration in home decorating: the gap between imagination and reality. We've all spent hours scrolling through furniture websites, trying to guess how a certain sofa or table would look in our own space. This process is filled with uncertainty, leads to buyer's remorse, and often requires a professional's eye to get right.

We wanted to create a tool that empowers anyone to become their own interior designer. Our vision was to leverage the power of generative AI to not just suggest products, but to create a photorealistic preview of a finished room, eliminating the guesswork and making the design process visual, fun, and risk-free.

What it does

DECORLY is an AI-powered interior design platform that transforms your ideas into photorealistic images of your dream space. It bridges the gap between browsing for furniture and seeing it in a fully decorated room.

The user journey is simple and intuitive:

  1. Define Your Space: Users start by selecting a room type (e.g., Bedroom, Living Room) and choosing their desired interior design theme (e.g., Minimalist, Bohemian, Modern). They can also specify a color palette to match their vision.
  2. Get AI-Curated Recommendations: Our AI analyzes these preferences and instantly suggests a curated list of real, shoppable furniture and decor items that fit the selected style.
  3. Select Your Products: Users can browse the recommendations and select the individual pieces they love for their room.
  4. Visualize the Magic: With a single click, DECORLY's AI takes the selected products and generates a stunning, high-quality, photorealistic image of a fully furnished room, showing exactly how the chosen items look together in a cohesive space.

Essentially, DECORLY lets you see it before you buy it, giving you the confidence to create a space you'll truly love.

How we built it

We built DECORLY using a modern, efficient, and scalable tech stack designed for a rich user experience and powerful AI integration.

  • Frontend: The user interface was built with Bolt, allowing us to create a clean, responsive, and intuitive design studio where users can seamlessly make their selections. The entire frontend is deployed on Netlify, which provides us with continuous integration, fast global delivery, and a robust hosting platform.

  • Backend & Database: We chose Supabase as our all-in-one backend-as-a-service. Supabase handles all our needs for user authentication, database management (for storing user-created spaces, preferences, and credit information), and storage, allowing us to build a secure and scalable application with minimal backend overhead.

  • AI: A lightweight python backend that leverages the OpenAI library. This service takes simple inputs like room type, theme, and with the help of simple prompts it chooses the products, then generate high-quality room images using OpenAI

Challenges we ran into

  • Prompt Engineering: Our biggest challenge was mastering prompt engineering. Crafting prompts to generate beautiful ui elements was not easy part. Had to refine the prompt precisely and then give it bolt to get stunning pages.

  • System Integration: Seamlessly connecting the Bolt frontend, Supabase for user data, and the Netlify offers only 10 secs timeout so had to use functions.

  • Balancing Simplicity and Control: Designing a UI that was simple enough for a beginner to use, yet offered enough control for a user to feel creative, was a delicate balance. We worked hard to make the product selection process feel empowering, not overwhelming.

Accomplishments that we're proud of

  • The "Magic Moment": We are incredibly proud of the core feature itself. The first time we saw a beautiful, coherent room generated from just a few clicks and text inputs was a huge "wow" moment and validated our entire concept.

  • A Seamless End-to-End Flow: We successfully built a complete user journey, from a simple sign-up to generating a stunning final product. The integration feels smooth, and the application delivers on its promise instantly.

  • A Lean and Powerful Architecture: By using a modern stack like Supabase and Netlify, we built a highly functional and scalable SaaS application with a very lean backend, allowing us to focus our efforts on the AI and the user experience.

What we learned

  • The Power of Prompting over Code: We learned that in the age of generative AI, the most impactful work often lies not in writing complex algorithms but in the art and science of communicating with the AI. The quality of our product is a direct result of creative and precise prompt engineering.

  • Modern Tools Accelerate Development: Leveraging platforms like Supabase and Netlify drastically cut down our development time. They handled the heavy lifting of authentication, database management, and deployment, so we could innovate on our core idea.

  • Visual Feedback is King: Seeing the positive reaction to the photorealistic renders reinforced our belief that for products like this, showing is infinitely more powerful than telling.

What's next for Decorly

We are just getting started and have an exciting roadmap ahead to make DECORLY even more powerful and indispensable for home decorators.

  • "Your Room" Mode: Implement functionality for users to upload a photo of their own room and use AI inpainting to realistically add, remove, or replace furniture within their actual space.
  • Direct E-commerce Integration: Partner with furniture brands to include "Add to Cart" functionality directly from the design studio, creating a seamless path from inspiration to purchase.
  • Style and Brand Expansion: Continuously add new design themes, color palettes, and a wider variety of furniture and decor items from more brands.
  • Enhanced Sharing & Community: Currently the generated images are not accurate, will be enhancing the prompt to get accurate results.

Built With

Share this project:

Updates