The idea: I wanted lyrics while working/coding, but existing solutions were either slow or locked to specific platforms.
Technical approach: - Built in C++ with raylib for performance (entire app is <12MB) - Uses Windows Media Session API for instant song detection (no audio processing needed) - Custom rendering library for the UI (gradient compositing, layer masking, FXAA)
Business model journey: - Started with $4.99/month subscription - Early feedback: "too expensive + subscription fatigue" - Pivoted to $19.99 one-time purchase - 100 song free trial, then lifetime access
Current state: - Windows only (macOS audio loopback is trickier, but planned using the new CoreAudio API) - Works with Spotify, YouTube, Apple Music, any app supporting Media Session - Just launched today, would love feedback
Happy to answer questions about the tech stack, business decisions, or anything else!