Inspiration

Sound Pilot started as my dream to make DJing truly accessible. I wanted to give anyone - whether beginner or pro - the tools to mix, perform, and connect with an audience in real time. The idea was simple; empower creators, build community, and let AI handle the hard parts so DJs can focus on music and interaction.

What it does

Sound Pilot is a live streaming DJ platform built for creators like me.

  • Dual decks: Mix tracks live with real-time waveforms and precise EQ.
  • Go live: Instantly broadcast my set to a global audience using LiveKit for high-quality, low-latency streaming.
  • Chat-driven control: I use natural language - typed or spoken to load tracks, trigger effects, or blend stems; the AI makes it happen.
  • AI SFX on demand: Describe a sound and ElevenLabs generates it; I drag and drop it straight into my mix.
  • Multi-stem mixing: Split tracks into vocals, drums, and more for creative live control.
  • Audience interaction: Viewers chat, request songs, and send reactions while I play.
  • Flexible interface: Collapsible panels let me focus on music or chat, depending on the vibe.
  • Subscriptions and upgrades: RevCat powers seamless in-app purchases and subscriptions for exclusive features.

How I built it

React SPA: Everything runs smoothly in the browser. Zustand: Manages complex audio and state logic efficiently. Tone.js: Handles playback, live processing, and waveforms. LiveKit: Delivers real-time, high-quality video and audio streaming for every show. RevCat: Manages subscriptions and upgrades, making it easy for users to support my work and unlock Pro or Autopilot features. OpenAI GPT-4o-mini: Powers the chat-based natural language controls. ElevenLabs API: Instantly creates AI sound effects on request. Tailwind CSS: Keeps the look modern and fully responsive. Lucide React: Provides clean, intuitive icons. Native File API: Lets me upload tracks and folders with a simple drag-and-drop.

Challenges I ran into

  • Low-latency audio and video is a must; LiveKit was key but needed fine-tuning.
  • Multi-stem playback and AI SFX generation had to work in sync without lag.
  • Mapping unpredictable chat commands to exact DJ actions pushed the limits of GPT-4o integration.
  • RevCat subscription logic needed to be rock-solid, especially for live shows with many users.
  • Balancing powerful features with a user-friendly interface took many iterations.

Accomplishments I’m proud of

  • Hosting live DJ sets directly from the browser with zero hardware required.
  • Enabling real-time audience chat, requests, and reactions making every show interactive.
  • Integrating seamless in-app purchases with RevCat so creators can monetize their talent.
  • Bringing AI-powered sound design and mixing into the live workflow.
  • Building a space where any DJ can find their crowd and grow their following.

What I learned

  • Real-time streaming and audio is always a challenge; tools like LiveKit and Tone.js are powerful when optimized.
  • External API integration (OpenAI, ElevenLabs, RevCat) means designing for both scale and failover.
  • Creators want simplicity and reliability never make them choose between features and flow.
  • Community and audience features drive engagement; every chat, request, and tip counts.

What’s next

  • Smarter AI DJ tools for intelligent transitions and auto-matching tracks.
  • New ways for fans to interact live polls, emotes, track voting.
  • More monetization options with RevCat; tipping, subscriptions, exclusive content.
  • Track history, hot cues, mobile streaming, and more FX.
  • Making Sound Pilot the Twitch for DJs, where every set can be a show.

Built With

Share this project:

Updates