GYST | Devpost

Inspiration

We wanted a music app that actually meets you where you are. Between school, part-time work, and life, our moods swing a lot. We asked: what if a playlist could gently steer you toward a steady “okay, I’ve got this” state instead of just matching your current vibe? That became GYST - Get Your Sh*t Together

What It Does

GYST estimates your current mood from a handful of lightweight signals (sleep/step trends, screen time, weather, calendar load, brief check-in). Then it builds a playlist curve, starting near your mood and nudging you toward neutral. At the end, you leave a simple emoji rating so the system learns what actually helped. Over time, GYST personalizes both the starting point and the curve.

How We Built It

Frontend: React + Vite for speed; Tailwind for a crisp, tactile UI.
State & logic: A lightweight scoring function ranks candidate tracks/sets along a 2D valence–energy model, then applies signal weights and your match/shift intent.
Architecture: Modular “connectors” pattern for future integrations (music providers, finance/health/weather context) with a clean boundary between UI, scoring, and data sources.

What's Next for GYST

Music provider integration: Spotify/Apple Music for live playlist creation and playback.
Context connectors: Optional finance (e.g., spend spikes), health, and weather inputs to auto-seed signals.
Personalization: Learn from reactions to fine-tune valence/energy and track selection over time.
Routine presets: One-tap “Morning Focus,” “Commute Calm,” “Pre-Gym Hype,” etc.
Privacy controls: Clear, opt-in data sharing with per-signal permissions.