Dario Miličić

New app build · Senior Flutter Engineer · Feb 2026 – Aug 2026

A music event booking platform, frontend built solo in Flutter web

Built the whole Flutter web frontend for booking artists, venues, instruments and equipment, from Figma to production, AI-assisted end to end.

~6 months
to production, against 18+ estimated without AI
44.8k
lines of Dart
560
automated tests
94
pull requests merged
Mousike: the landing page with the catalogue of venues, artists and instruments

Challenge

Mousike lets event organisers plan a music event in one place: book artists and a venue, add instruments and equipment, and pay at checkout. It needed a production web app built from Figma designs, against a Django REST API a colleague was building at the same time.

What I did

  • Built the frontend end to end: 14 feature areas, including a six-step booking flow (details, artists, venue, instruments, equipment, checkout), the cart, invoices, availability calendars and management screens for artists, venues and equipment.
  • Structured every feature in Clean Architecture layers, with Riverpod for state and dependency injection, GoRouter for auth-aware routing, and a typed Result instead of exceptions between layers.
  • Implemented the Figma designs for desktop, tablet and mobile web from one codebase, on a shared Mousike component library.
  • Wrote the rules the AI agents follow (CLAUDE.md, ARCHITECTURE.md and a custom responsive-design skill), then built with Google Antigravity and, from May, Claude Code.
  • Shipped through CI that builds the WebAssembly release and deploys every merge to AWS S3 and CloudFront.

Outcome

  • In production at dashboard.mousike-app.com.
  • 44.8k lines of Dart across 14 feature areas, covered by 560 automated tests.
  • Shipped in about six months with AI, against an estimated 18 months or more for the same scope without it.

Stack

  • Flutter
  • Dart
  • Flutter Web
  • WebAssembly
  • Riverpod
  • GoRouter
  • dart_mappable
  • Clean Architecture
  • REST
  • GitHub Actions
  • AWS S3
  • CloudFront

How it was built

This project is the worked example for my AI workflow: the tools and models used, and the numbers before and after. See how I build with AI →

Links