I developed HearthBeats, a Python and Streamlit program for analyzing tenant feedback in collaboration with North Star Housing Group. The platform processes structured and unstructured survey data using NLP techniques such as sentiment analysis, topic modeling, and emotion detection.
It features a Feature Variability Index to measure feedback stability, Geo-Aware Analysis with interactive maps to uncover location-based insights, and an Invisible Voices module to highlight underrepresented tenant perspectives.
Built with Pandas, NumPy, spaCy, scikit-learn, PyTorch, Sentence-Transformers, Plotly, Seaborn, and Folium, the program delivers interactive dashboards, spatial analytics, and bias-aware outputs, transforming raw tenant data into actionable, equitable insights for housing stakeholders.