Mainstage

Inspiration

As a DJ, I found that preparing for a performance often takes longer than performing it. Modern DJ software is excellent at organising music, but it still expects DJs to think in folders, playlists and metadata. That's not how most DJs prepare.

We think in intent.

"I need a warm-up set that grows naturally."

"I don't want to outshine the headliner."

"I need a bridge between funk and breaks."

Mainstage was born from a simple question:

What if AI understood musical intent instead of just organising tracks?

Rather than replacing the DJ, I wanted to build an assistant that helps transform creative ideas into a structured performance plan while keeping the DJ firmly in control.

What I'm Building

Mainstage is an AI-powered preparation environment for DJs.

It connects to a DJ's existing music ecosystem, understands the event they're preparing for, analyses the music they already own (and, where supported, connected streaming services), and produces an explainable performance plan that the DJ can review, refine and export back into their existing workflow.

The goal is not to generate playlists.

The goal is to help DJs prepare better performances.

How I Built It

Before writing code, I focused on product design.

I defined three foundations:

  • A product architecture built around connector-based music sources rather than a single DJ platform.
  • A local-first desktop architecture designed to work alongside existing DJ software.
  • An AI behaviour specification that defines how Mainstage reasons about performances, ensuring recommendations are collaborative, explainable and always leave creative control with the DJ.

By separating product design, architecture and AI behaviour first, implementation can stay focused without losing sight of the original vision.

Challenges

The biggest challenge has been deciding what not to build.

As a DJ, there are hundreds of features that would be useful, but Build Week forced me to focus on a single question:

What is the smallest product that demonstrates something genuinely new?

That led to a much clearer vision. Mainstage isn't trying to replace Rekordbox, Serato or Traktor. It complements them by solving a different problem: preparation.

Another challenge has been designing AI that collaborates rather than dictates. Every recommendation should be explainable, every decision reversible, and the DJ should always remain in control.

What I Learned

The biggest lesson has been that AI becomes far more valuable when it understands intent rather than simply automating tasks.

For DJs, preparing a performance is a creative process. Mainstage isn't trying to automate creativity. It's designed to support it by helping DJs organise ideas, explore possibilities and prepare with greater confidence.

What's Next

The next milestone is bringing the design to life: integrating real music libraries, connecting with DJ software, refining the AI reasoning engine and validating the experience with working DJs.

My long-term goal is to build a tool that feels less like an algorithm and more like an experienced resident DJ helping you prepare for your next performance.

Built With

Share this project:

Updates