Synopsis
TYPE
mobile
DURATION
1 month
COMPLETED
March 2025
LANGUAGE
Dart
MISSION BRIEF
Synopsis is a modern music streaming platform built with Flutter and Supabase, designed to deliver a fast, intelligent, and immersive listening experience. The app combines real-time synced lyrics, powerful offline support, and a highly adaptive recommendation engine that improves the more you listen. With features like smart caching, predictive prefetching, and an AI DJ that learns user taste over time, Synopsis offers a personalized music journey that feels alive and intuitive across devices.
VISUAL_EVIDENCE
CORE FEATURES
- High-quality music streaming
- Synced lyrics with real-time playback highlighting
- Album and artist support
- Follow artists and receive updates
- Create and manage playlists
- Like and save favorite songs
- Smart caching for faster repeat playback
- Offline song downloads
- Prefetching the next track for seamless transitions
- AI DJ that adapts to listening patterns
- Personalized music recommendations
- Recently played and listening history
- Cross-session preference learning
- Low-data mode for limited connectivity
TECH STACK
- >Flutter for cross-platform mobile development
- >Supabase for authentication, database, and file storage
- >Row-level security (RLS) for user-specific data isolation
- >Optimized audio streaming with adaptive buffering
- >Local disk caching for frequently played tracks
- >Background audio playback support
- >Offline-first architecture with sync reconciliation
- >Prefetching engine for predictive track loading
- >AI-driven recommendation system using listening history and behavior signals
- >Real-time lyrics synchronization using timestamped lyric data
- >Scalable media storage and CDN-backed delivery
- >Clean architecture with modular state management
Operational Hurdles
Ensuring smooth playback under unstable network conditions
SOLUTION_PROTOCOL: Implemented adaptive buffering, intelligent caching, and fallback bitrate strategies to maintain uninterrupted playback
Designing a recommendation system that improves over time without explicit user input
SOLUTION_PROTOCOL: Built a behavior-driven learning model using listening duration, skips, replays, and likes to continuously refine recommendations
Managing offline downloads while keeping storage usage efficient
SOLUTION_PROTOCOL: Used encrypted local storage with smart eviction policies based on listening frequency and available disk space
Synchronizing lyrics accurately across different devices and playback speeds
SOLUTION_PROTOCOL: Leveraged timestamp-based lyric mapping with dynamic drift correction tied to the audio playback engine
Creating an AI DJ that feels natural rather than repetitive
SOLUTION_PROTOCOL: Introduced session-aware song sequencing, diversity constraints, and contextual awareness based on time and recent activity