Spotilike

Spotilike is an AI-powered music app that reads your mood through expressions to play songs that match how you feel. It learns from your skips and smiles to personalize your emotional soundtrack.

Comment

Inspiration

Spotilike was inspired by the desire to create a more emotionally intelligent music experience. Traditional recommendation systems rely on genre, artist, or listening history, but often miss the user's current mood. We set out to bridge that gap by using AI to understand real-time emotions and personalize music accordingly.

The project combines:

  • Emotion Detection: Real-time facial expression recognition
  • Music Intelligence: Spotify’s music database and recommendation engine
  • Real-time Learning: Adapting recommendations based on user reactions

What it does

Spotilike is a full-stack app that recommends music based on how you feel and what you’re going through.

Real-time Emotion Detection

  • Uses your webcam to detect emotions like happy, sad, angry, or surprised
  • Tracks how you react to songs
  • Learns which songs improve your mood

Context-Aware Recommendations

  • Users describe their situation (e.g., “I got a promotion” or “I’m overwhelmed”)
  • Google Gemini AI extracts the mood and topic from the text
  • Matches music to both your emotion and context

Smart Playback Control

  • Full Spotify playback control
  • Real-time progress tracking and controls
  • Displays album art and track info

Automatic Updates

  • Tracks and updates your emotional responses in real time
  • Uses Server-Sent Events for instant UI feedback

How we built it

Backend (Python/Flask)

  • Flask API: Handles routes, logic, and Spotify control
  • Spotipy: Integrates Spotify Web API
  • MongoDB: Stores emotions, song scores, and user reactions
  • DeepFace: Detects facial emotion
  • Google Gemini: Analyzes text for mood and topic

Frontend (Next.js/React)

  • Next.js 14 with TypeScript
  • Tailwind CSS for styling
  • Responsive UI with live data using SSE
  • Works on both desktop and mobile

Key Features

  • Secure Spotify OAuth
  • Real-time emotion detection via webcam
  • Resilient error handling and fallback logic
  • API response caching for performance

Challenges we ran into

Technical

Emotion Detection

  • Balancing speed and accuracy
  • Handling lighting and webcam quality
  • Managing browser permissions and CPU usage

Spotify API

  • Handling complex OAuth flow
  • Keeping playback state in sync

Database

  • Designing schema to score songs by emotion
  • Real-time writes and reads for fast feedback

Real-time Communication

  • Implementing SSE and handling reconnects
  • Keeping frontend and backend in sync

UI/UX

  • Making emotion detection feel natural
  • Designing intuitive music controls
  • Responsive layout across devices
  • Balancing functionality with simplicity

Accomplishments that we're proud of

Technical

AI Integration

  • Combined DeepFace and Gemini into one system

Spotify Playback

  • Full Spotify control with real-time sync
  • Built a modern, responsive music player

Real-time Feedback

  • UI updates instantly based on detected emotion
  • SSE for smooth and efficient updates

Error Resilience

  • Graceful fallback systems for AI and API errors
  • Clear error messages and auto-recovery

User Experience

Clean Interface

  • Spotify-inspired design
  • Mobile-friendly and accessible

Emotional Intelligence

  • Learns from how users react
  • Improves over time with context-based recommendations

Accessibility

  • Keyboard navigation and screen reader support
  • High contrast, visible feedback

What we learned

AI

  • Emotion detection is powerful but sensitive to lighting and expression
  • Multiple AI services require good error handling
  • Real-time processing needs performance optimization

Music Recommendation

  • Emotions deeply influence music preferences
  • Facial feedback gives useful, real-time user data
  • Spotify API is robust but requires careful handling

Full-Stack Dev

  • SSE is a great alternative to WebSockets
  • Managing state across frontend/backend is challenging
  • Error handling is critical for a smooth experience

UX

  • Emotionally responsive design boosts engagement
  • Performance is key for music apps
  • Accessibility is easiest when built from the start

What's next for Spotilike

Short-term (Next 3 months)

Better Emotion Detection

  • Detect subtle moods
  • Support for biometric data
  • Improve webcam reliability

Deeper Music Understanding

  • Analyze lyrics and song energy
  • Use emotional metadata from music APIs

Social Features

  • Share emotion-based playlists
  • Mood-based collaborative sessions
  • See what friends with similar moods are listening to

Mid-term (3–6 months)

Smarter AI

  • Train models to personalize mood detection
  • Use behavioral data for better recommendations
  • Predict mood changes and suggest music in advance

Platform Expansion

  • Launch mobile apps
  • Build a desktop app
  • Integrate with smart speakers

Analytics

  • Track mood over time
  • Correlate emotional trends with listening habits
  • Offer insights and playlists based on mood history

Long-term Vision (6+ months)

AI Music Creation

  • Generate music based on your emotional profile
  • Adjust playback in real-time as your mood shifts
  • Collaborate with AI to create music

Ecosystem Integration

  • Connect to smart home devices and wearables
  • Work with health apps for wellness tracking

R&D and Community

  • Partner with researchers on emotion-music studies
  • Share findings and open-source our tools

Tech Roadmap

Performance

  • Move emotion detection to edge devices
  • Optimize AI models for speed
  • Cache music data efficiently

Scalability

  • Shift to microservices
  • Cloud deployment with auto-scaling
  • Global CDN support

Security & Privacy

  • End-to-end encryption
  • Local-only emotion detection options
  • Full GDPR and privacy compliance

Spotilike is building the future of emotionally aware music — where your playlist truly understands how you feel.

Built With

Share this project:

Updates