LyricLens

LyricLens is an AI-powered song lyrics analysis tool. It help users uncover the hidden depths of their favorite songs by providing a comprehensive breakdown of themes, emotions, literary devices, and structure.

Features

• Deep Lyrics Analysis: Uses advanced AI (gemma-4-26b-a4b-it) to deconstruct song meaning.
• Visual Metrics (Radar Chart): Visualizes 8 key dimensions of a song:
  • Emotion, complexity, storytelling, imagery, repetition, intensity, originality, and depth.
• Multilingual Support: Generate analyses in English, German, Spanish, French, Italian, or Japanese.
• File Upload: Supports pasting lyrics directly or uploading .txt files.
• Analysis History: Keeps track of your recent analyses (stored on a PHP/MySQL backend).
• Responsive & Modern UI: Built with React, Tailwind CSS, and Framer Motion for a smooth, immersive experience.

Tech Stack

• Frontend: React, Vite, Tailwind CSS, Recharts, Framer Motion, Lucide Icons.
• Backend: PHP (for API), MySQL (for history storage).
• AI Integration: Openrouter or Ollama

Demo

https://lyriclens.jentsch.io/

Getting Started

Prerequisites

• A web server (Apache/Nginx) with PHP and MySQL support.
• Node.js & npm (for building the frontend).
• A running Ollama instance (optional, for local AI).

Backend Setup

1. Create a MySQL database named lyriclens.
2. Run the SQL schema found in backend/database.sql to create the history table.
3. Configure your database credentials in backend/api/config/config.php.
4. Upload the contents of the backend/ folder to your web server (e.g., to /api).

Frontend Setup

1. Clone the repository.
2. Install dependencies:

npm install

3. Update the config.ts file in the /src directory:

const config = {
  api: {
    generateUrl: "/api/generate.php",
    historyUrl: "/api/history.php",
  },
  model: {
    name: "google/gemma-4-26b-a4b-it",
  },
} as const;

export default config;

4. Build the project:

npm run build

5. Deploy the contents of the dist/ folder to your web server.

Development

To run the development server locally:

npm run dev

AI Studio "Vibe-Coding"

This project was built using "Vibe-Coding" methods in Google AI Studio. It showcases how a developer can iterate rapidly from a simple idea to a complex full-stack application by directing a coding agent through natural language.

---

Developed by Michael Jentsch | jentsch.io Copyright © 2026