MuseMatch

AI-powered mood and intent analysis using LLMs to understand user emotions and suggest relevant resources based on creative context.

Comment

MuseMatch

Inspiration

In a world where creative blocks and moments of feeling down are common, we wanted to build more than just a resource finder. Our inspiration was to create a digital companion that genuinely lifts spirits and sparks creativity. We envisioned an application that doesn't just give you links, but provides a multi-sensory, uplifting experience with a personal touch—using humor, custom visuals, and a friendly voice to bring a smile to anyone's face.

What it does

MuseMatch is an AI-powered web application designed to help users overcome creative ruts and negative feelings. Users describe their current mood or creative goals, and MuseMatch's AI engine generates a curated set of inspiring and humorous ideas.

  • AI-Powered Suggestions: Get unique ideas for articles, videos, memes, tutorials, and more, tailored to your input.
  • Visual Inspiration: Each idea is paired with a unique, AI-generated image to spark your imagination.
  • Uplifting Urdu Narrations: Every suggestion comes with a funny story or joke in Urdu. With the click of a button, you can hear it narrated in a sweet, female voice.
  • Engaging UI/UX: The app features an eye-catching design, beautiful animations, and background music to create a deeply engaging and comforting experience.

How we built it

MuseMatch is a modern web application built with a focus on performance, user experience, and powerful AI capabilities.

  • Frontend: Next.js, React, and TypeScript for a fast, type-safe, and scalable user interface.
  • Styling: We used Tailwind CSS with ShadCN UI components to create a polished, responsive, and aesthetically pleasing design.
  • Generative AI: The core of MuseMatch is powered by Google's Gemini models through the Genkit framework.
    • Idea & Image Generation: A Genkit flow processes the user's mood to generate a list of creative ideas and prompts the Gemini image model to create a unique visual for each one.
    • Text-to-Speech: Another flow uses Gemini's TTS model to convert the humorous Urdu descriptions into high-quality audio narration with a specified Urdu voice.

Challenges we ran into

  • API Rate Limiting: Initially, generating audio for all suggestions at once quickly exhausted the free-tier API quota, causing the app to fail. We solved this by implementing an on-demand "Narrate" button for each card, which generates audio only when the user requests it. This made the app more robust and efficient.
  • Performance: Generating multiple AI images and text descriptions simultaneously was slow. We optimized the process by running the AI generation tasks in parallel on the backend, significantly improving the perceived performance and reducing the waiting time for the user.
  • Voice and Language Support: Finding a text-to-speech voice that supported Urdu and fit the "sweet and funny" persona required experimentation. We had to debug "language not found" errors and select a compatible voice from the available options to deliver the desired experience.

Accomplishments that we're proud of

We are incredibly proud of creating an application that feels personal and alive. The combination of AI-generated text, images, and voice narration creates a truly unique and multi-sensory experience. Overcoming the technical hurdles to make the app responsive and reliable without sacrificing its core "magical" features was a major accomplishment. The result is a fun, engaging, and genuinely helpful tool.

What we learned

This project was a deep dive into the practical application of modern generative AI. We learned how to orchestrate multiple AI models (text, image, and audio) within a single user experience. We also learned the importance of managing API limitations gracefully and the significant impact of backend architecture on frontend performance. Most importantly, we learned that the best user experiences often come from combining technology with a strong sense of empathy and a touch of humor.

What's next for MuseMatch

The journey is just beginning! We have many exciting ideas for the future:

  • Video Generation: Explore emerging text-to-video models to generate short, animated clips for video ideas.
  • User Accounts & History: Allow users to create accounts to save their favorite ideas and track their creative journey.
  • More Languages & Voices: Expand our language support to bring the MuseMatch experience to a global audience.
  • Deeper Personalization: Fine-tune the AI to learn from user feedback and provide even more personalized and effective suggestions over time.

Built With

Share this project:

Updates