Developer creates MoodSense AI, a practical NLP application that detects emotional sentiment from text and delivers mood-based recommendations through both API and web interface.

MoodSense AI classifies text into multiple mood categories (happy, sad, anxious, etc.) with confidence scores and probability distributions

Built with Python, scikit-learn, LightGBM, and spaCy for the ML backend, with FastAPI for API exposure and Gradio for the user interface

Project deployed on Hugging Face Spaces with a live demo, emphasizing practical usability beyond model training alone

Includes both REST API (FastAPI) and interactive UI (Gradio) interfaces, making the mood detection accessible for different use cases