Case Study / Shipped SaaS

SoundSauce

Hear a sound. Leave with the preset.

Producers spend hours reverse-engineering sounds from YouTube tutorials and sample packs. SoundSauce collapses that: upload any audio clip and get back a free preset for the Vital synth that recreates the sound, plus a breakdown of how it was made. Built solo as a complete SaaS, from the DSP analysis to the Stripe billing.

  1. Uploadany clip, any source
  2. Separatestems via Replicate
  3. Detectinstruments via Gemini
  4. AnalyzeDSP in Web Audio
  5. Generatea Vital preset file
Under the hood

A whole product, not a demo.

The audio brain

  • Stem separation through Replicate so a full mix breaks into analyzable parts
  • Gemini classifies what instrument it's hearing before analysis starts
  • Web Audio API extracts the waveform, filter curve, ADSR envelope, and modulation
  • Output is a real .vital preset the producer opens in their synth

The SaaS around it

  • Supabase auth, storage, and Postgres with row-level security across 22 migrations
  • Stripe subscriptions gating a freemium Pro tier
  • A social layer: profiles, follows, likes, comments, and sound-design challenges
  • PostHog, Sentry, web-vitals, Playwright and Vitest suites, CI on GitHub Actions

The honest coda: SoundSauce shipped in spring 2026, had its launch moment on the Vital community's corner of Reddit, and settled into quiet. It stays live as a working product and as proof of range, the whole stack of a modern SaaS, built and operated by one person.