PCA Insights System

Healthcare authorities collect vast amounts of qualitative patient feedback, but extracting meaningful insights from thousands of comments is a significant challenge. The Patient Comment Analysis System (PCAS) was developed to help health authorities transform raw patient experience data into actionable intelligence—enabling administrators to identify trends, address concerns, and improve care quality across their facilities.

The system was engineered with healthcare workflows at its core. A clean, responsive dashboard allows staff to upload Excel or CSV files, automatically map columns from various data formats, and instantly receive comprehensive analysis. The interface prioritizes clarity and efficiency—sidebar filters let users drill down by facility, health authority, or theme, while visualizations present complex data in digestible formats. Privacy was paramount: all processing happens locally on the user's machine, with options for air-gapped operation to ensure PHI data never leaves the facility.

Built with Python and Streamlit, the system leverages natural language processing through spaCy for theme detection, combined with custom sentiment analysis and actionability scoring algorithms. Performance optimizations include database caching, lazy loading of ML models, parallel processing for large datasets, and SQLite with WAL mode for efficient data management. The modular architecture ensures maintainability while the standalone executable packaging makes deployment straightforward for non-technical healthcare staff.