Music Meets Alexa

Meet the person behind the playlist: a simulated Alexa+ experience that turns shared music tastes into conversation starters, with consent-first connections.

Comment

Inspiration

What if music could introduce us to people, not just more music?

A favorite song can give us something more personal to talk about than a job title or a profile photo. Music Meets Alexa explores that idea: listen to a person’s musical world, discover common ground, and decide whether to take the next step.

The project builds on Coffee Meets Codex, adapting its intention-based discovery and consent-first approach into a music-led experience.

What it does

Music Meets Alexa is a simulated Alexa+ web experience featuring fictional profiles. Users choose a Friend or Lover intention, hear short introductions, and explore each person’s musical identity.

  • Meet: discover people through a daily listening brief.
  • Hear: listen to ten original music loops synthesized in the browser.
  • Understand: see explanations grounded in shared taste information.
  • Start a conversation: explore a three-track mixtape and a suggested opener.
  • Choose: opt in through two confirmations before reaching Pending.

No real person is contacted, and Pending does not represent a completed introduction.

How we built it

We built the interface with JavaScript, HTML, and CSS, alongside a Node.js MCP server exposing six tools.

Amazon Polly generated prerecorded Danielle voice responses, played as MP3 files with browser speech synthesis as a fallback. This is our simulation’s voice, not an official Alexa device voice integration.

The Web Audio API synthesizes ten original music loops live in the browser. Track selection and the Now Playing display use the same song data so that what users see matches what they hear.

Match explanations and shared-taste labels also use the same underlying overlap data, keeping the spoken and visual information consistent.

The demo is deployed on Render, and the source is available on GitHub.

Challenges we ran into

Keeping sound and screen in sync

Each selected track needed to play its own loop, with the Now Playing display reflecting the actual playback selection. We added checks for the track-to-loop mapping.

Making explanations consistent

An early version could describe different tastes while displaying a shared term. We corrected this by calculating the overlap once and using it for both the explanation and the shared-taste display.

Balancing voice, detail, and consent

Spoken introductions need to be brief enough to follow, while fuller details remain available on screen. Expressing interest also needed to be an explicit choice, supported by confirmations and a visible Pending state.

What we learned

Voice, music, and visual information need different pacing. A short spoken teaser invites curiosity; the screen gives users time to explore.

We also learned that small inconsistencies can weaken trust quickly. A mismatched song or contradictory explanation can interrupt an otherwise coherent experience. Testing the connections between features mattered as much as building the features themselves.

Music works best here as a starting point for conversation, not as a claim that an algorithm can determine compatibility.

Accomplishments we're proud of

We brought spoken introductions, original browser-synthesized music, grounded explanations, and consent controls into one working experience that people can try directly.

What's next

We want to test whether music-led introductions help people start meaningful conversations, improve accessibility and voice pacing, and explore a real Alexa+ integration.

A version involving real people would require identity verification, reciprocal consent, and appropriate privacy controls before enabling introductions.

Attribution and prototype scope

This project adapts the intention model, fictional cast, and consent-first patterns from Coffee Meets Codex, adding the music experience, MCP server, and browser simulator.

This is a simulated Alexa+ experience, not a deployed Alexa device integration. Profiles are fictional, and no real introductions are made.

Built With

Share this project:

Updates

Submission history