• 

  Home page

• 

  Result page

MuseAI

Inspiration

Art has always fascinated me. I love how museums connect people to history, culture, and creativity. But not everyone can travel to visit the world’s great museums, and sometimes, even when you’re there, it’s hard to find clear explanations or to feel engaged. That’s why I wanted to build MuseAI — a tool that makes art more accessible and interactive for everyone, anywhere in the world.

What I Learned

While building MuseAI, I learned:

• How to integrate multiple APIs and manage async processes for fast responses.
• How to generate and stream audio dynamically with ElevenLabs.
• How to deploy a modern web app quickly using Netlify and manage custom domains with Entri and IONOS.
• How to store user data securely with Supabase.
• How to design a clean, minimal UI that works well on mobile and desktop.

How I Built It

MuseAI is a web application built from scratch using:

• Bolt.new for backend logic and routing.
• OpenAI GPT-4 to generate smart, engaging text about artworks.
• ElevenLabs to turn that text into lifelike voice narrations.
• Supabase to store user history, allowing users to review their past scans and keep learning.
• Netlify to deploy the app to the custom domain muse-ai.site.
• Frontend technologies including React, Tailwind CSS, and animations with Framer Motion.

The user can:

1. Scan or upload a picture of an artwork.
2. Get an instant AI-generated description about the piece.
3. Listen to that description through a realistic voice narration.
4. Ask follow-up questions to dive deeper into the art.
5. Save their scan history for future reference.

Challenges I Faced

Building MuseAI came with several challenges:

• Managing multiple API calls in parallel to keep the experience smooth and fast.
• Handling audio generation delays and synchronizing the playback with user actions.
• Designing an interface that feels simple yet powerful, without overwhelming the user.
• Deploying and managing a custom domain to give the app a professional look and feel.
• Keeping latency low, especially for voice streaming and text generation.

Despite these challenges, I’m proud of how MuseAI turned out, and I believe it can help make art more accessible and enjoyable for people everywhere.

Try It Out!

Check it out here: https://muse-ai.site

Built With

• bolt.new
• elevenlabs
• framer-motion
• html
• javascript
• netlify
• openai
• react
• supabase
• tailwindcss