Project Lead: Abraham D. Flaxman, UW Global Health

Data Science Lead: Bryna Hazelton, Naomi Alterman

The purpose of this project is to modernize SmartVA-Analyze, an existing open-source verbal autopsy (VA) software, by transitioning its codebase from Python 2 to Python 3. Verbal autopsy is a crucial tool for estimating the causes of death in settings where medical certification is unavailable, particularly in low- and middle-income countries. By collecting data from interviews with the deceased’s relatives or caregivers, VA provides insights into mortality patterns, supporting the formulation of health policies and guiding the allocation of resources. However, with Python 2 being discontinued, maintaining and deploying the current software has become increasingly challenging, limiting its accessibility and functionality.

The proposed update aims to ensure compatibility with modern computing environments, enhance the software’s usability, and enable easy installations. By doing so, we not only maintain the software’s current functionality but also position it for integration with advanced artificial intelligence (AI) tools, specifically large language models. These future enhancements will expand the software’s analytical capabilities, potentially automating parts of the VA process and enabling richer insights from collected data.