The problem:
Existing concert finders like Songkick and Bandsintown are nice if you know exactly where and when you want to see a show. But they fall short for flexible planning. I wanted a tool to answer questions like, "I want to take a road trip from CA to TX sometime this summer, are there any shows worth seeing along the way?"
The solution:
Concert Map lets you explore all upcoming shows on a map. You can filter by date, genre, or search for individual artists. Events are scored to show you the biggest artists playing in any given region, but you can always zoom in to see more.
You can personalize the map by importing your top played artists from Last.fm or Spotify. It then highlights events according to your listening history. And for music festivals, it shows lineups with your top artists highlighted, so you can immediately see which festivals are worth the trip. The About page [1] has screenshots showing what this looks like in action.
The tech stack:
- backend: Rust with Axum
- frontend: Svelte 5 with shadcn-svelte
- marker mapping: Leaflet.js
- data sources: ListenBrainz, Last.fm, Jambase, Ticketmaster
- deployment: Render (postgres, cron jobs, and web service instances)
- The original data ingestion pipeline used several GB of peak memory, which was too much for a cheap cron job instance on Render. I rewrote it to stream + process the json data chunk by chunk, which dropped peak memory usage to under 100MB, allowing it to run on the cheapest tier. Rust makes this surprisingly straightforward.
- Processing and rendering hundreds of thousands of clustered event markers. Added Web Workers to parallelize this so the map loads in a reasonable amount of time. Still plenty of room for improvement here.
- User accounts for syncing saved artists across device
- Integrating more music service and event data providers
- Save individual events to a list