Playlist Generator
From a written vibe to a curated playlist. The app was designed and built end to end with AI development tools—Figma Make, V0, GitHub Copilot, and Cursor—then shipped on Vercel.
The exploration
How far an AI-assisted workflow can take a product from first design pass to a live Next.js deploy, while playlist matching still uses intentional, hard-coded emotion rules.
Built with
Figma Make / V0 / GitHub Copilot / Cursor / React / Next.js / Vercel
The challenge
Explore AI-assisted development by shipping a real product quickly: analyze free-form user input, map it to mood and genre, and return a playlist—going from first design pass to a public deploy in a short window.
The approach
I used Figma Make for the initial design and flow, V0 for UI scaffolding, and GitHub Copilot with Cursor to implement the React and Next.js app. Input analysis uses hard-coded emotion mappings rather than a live model, which kept the logic clear and easy to evolve. Vercel handled hosting with automatic deploys from the repository. The full path—from design to live—took about two hours.
The outcome
The live app generates playlists from written input and shows how AI tools can compress a design-to-deploy cycle without skipping craft. The emotion layer is intentionally simple so richer analysis can replace it later; the experiment’s value is the end-to-end workflow as much as the feature itself.