Sonic Cartographer

Sonic Cartographer uses AI to turn your listening history into a musical identity portrait. It converts passive consumption into active exploration to expand your tastes authentically.

Comment

Inspiration

I have always loved music, starting in my late teens / early 20s where I spent hours every day listening to music, reading about it in magazines and books as well as roaming my local record shops to find the next artists/bands to listen to...

I have always tried to understand who I was as a music fan, what unknown artists and genres I could explore next... I wanted to understand and demonstrate how I could leverage AI to help with music discovery...

What it does

Sonic Cartographer addresses a fundamental challenge in music discovery: helping listeners understand and expand their musical identity while maintaining authentic connection to their tastes.

The application transforms raw listening history into actionable insight through AI-powered analysis, converting passive consumption patterns into active exploration strategies.

Users upload their artist collections or recent listening history, receive detailed listening portraits that reveal hidden patterns in their preferences, engage in guided conversations that clarify discovery goals, and receive curated album recommendations that balance familiarity with intentional expansion.

How we built it

I used the following workflow to build the app:

  1. Brainstorming & ideation using Miro and conversations with an LLM (Gemini)
  2. Design and front-End development of a full MVP end-to-end user journey using Figma Make
  3. Back-End requirements, design, scaffolding and deployment using VSCode + Claude Code + Raindrop MCP via /new-raindrop-app command
  4. Incremental implementation and testing of the back-end API endpoints and front-end/back-end integration using VSCode + Claude Code

Challenges we ran into

  1. Time - I was only able to start 10 days before submission days and was only able to spend a limited amount to time on this initially
  2. Understanding of the new workflow - it took me a while to understand the recommended workflow
  3. Raindrop /new-raindrop-app command felt like a blackbox - where running the command, I did not feel I had a good understanding of what was actually going on
  4. Learning new tech stack at speed - I did not have the time I felt I needed to fully understand and test what Raindrop and Vultr offered and I definitively cut some corners (for example, I did not spend enough time digesting the documentation or going through the tutorials)

Accomplishments that we're proud of

  1. Building a working app with in a few days with a tech and tooling stack I never used before 2.Finding the time to fully participate to a hackathon for the first time in years
  2. Experiment with new tech and new workflow

What we learned

  1. When I face new tech and new workflow, I should not rush into implementing my idea and instead I should try first to go through tutorials and simple experiments to make that I have a good grab on tech and workflow
  2. I knew this already, but LLM and coding assistants needs very clear instructions and conversation, to produce good quality outputs at speed

What's next for Sonic Cartographer

  1. User research/testing - show what I have produced to friend and colleagues to see if this solves an actual problem for others and get their view on feature and user journey
  2. Finish the MVP - including bug/performance fixes, LLM tuning and queries optimisation, full end-to-end user journey, Google Auth, etc.
  3. Add new features (if validated by user research) - Spotify/Discogs integration, voice conversation, mobile app front-end , etc.

Built With

Share this project:

Updates