Team Manager

Personal product.

A web and mobile app that centralises team and sports activity management, with AI (Gemini) that helps plan trainings, prepare matches and analyse team information.

Clubs and coaches split rosters, matches, trainings and stats across spreadsheets and chat apps. Analysis arrives late, or never.

• Product definition and architecture.
• Web client in Next.js and mobile client in React Native (Expo) on the same backend.
• Supabase database with schemas as the source of truth and generated migrations.
• Frontend and database tests, plus continuous integration.
• Gemini integrated into trainings, programmes, matches and the medical area, always as a reviewable suggestion.
• Automated mobile releases: tagging a version makes GitHub Actions trigger the Expo EAS builds.

1. ASQL schemas as the single source of truth: migrations are generated by diff, never by hand, so schema and migrations cannot drift.
2. BDatabase contracts tested with pgTAP, alongside UI tests with Vitest and Testing Library.
3. CA shared core for club operations and sport-specific models for statistics and game logic.
4. DWeb/mobile parity: every significant change is checked in both clients.

A workspace built on rosters, players, matches, trainings and statistics, with AI as an assistant that suggests rather than something that decides for the coach.

Key features

• Rosters and player profiles
• Matches and trainings
• Statistics and comparisons
• AI assistance
• Web and mobile app
• Light and dark mode

• Next.js
• React
• TypeScript
• Supabase
• PostgreSQL
• React Native
• Expo EAS
• Gemini API
• Tailwind CSS
• Docker
• Vitest
• pgTAP
• GitHub Actions

Private project: there is no public demo or open repository.