🎶 BeatScouter

BeatScouter is an Android app that recommends songs similar to a given track using a custom-built Flask + Machine Learning API.
It helps users discover music based on cosine similarity of Spotify audio features.

---

📱 App Features

• 🔍 Search for a track and artist
• 🎵 Get personalized song recommendations from the backend API
• 🖼️ Display album artwork, track name, and artist name
• 📊 Show similarity score for each recommendation
• 🎧 Tap track URLs to open directly in Spotify

---

⚙️ How It Works

1. User enters a track name and artist in the app.
2. The app sends a POST request to the Flask API endpoint /recommend.
3. The API finds the most similar songs based on cosine similarity across normalized Spotify audio features.
4. The results (including track_name, artist_name, track_url, artwork_url, and similarity_score) are returned as JSON.
5. BeatScouter displays them in a RecyclerView with a clean UI.

---

📡 Tech Stack

🔹 Frontend (Android App)

• Kotlin
• RecyclerView for track display
• Retrofit for API requests
• Gson for JSON parsing
• Glide/Picasso for album artwork loading

🔹 Backend (API)

• Flask (Python)
• Pandas, NumPy, Scikit-Learn for feature scaling & similarity
• Cosine similarity on Spotify audio features
• Returns results in JSON format

---

📂 Project Structure