Purpose:
Built to explore how Natural Language Processing (NLP) and semantic search can translate human emotions into personalized music experiences, demonstrating practical AI applications beyond traditional analytics.

Process and Technology:

• Users enter free-text descriptions of their mood

• A SentenceTransformer model converts text into vector embeddings

• Semantic similarity is calculated against pre-encoded playlist descriptions

• The app identifies and recommends the most relevant playlist with a Spotify link

• Built with Python, Streamlit, and deployed via Streamlit Cloud for public access

ELT Perspective:

• Extract: User mood input captured via the web interface

• Load: Data transformed into embeddings using NLP models

• Transform: Vector comparisons generate similarity scores, delivering real-time recommendations

• Minimal data storage requirements due to lightweight in-memory architecture

This project highlights skills in NLP, semantic search, real-time data transformation, and end-to-end deployment of an AI-driven user experience.