SUMMARY Financial incentives are an evidence-based intervention to improve engagement in beneficial health behaviors and services for many priority health issues. Our team has extensively studied the benefits of financial incentives for people living with HIV in Tanzania, including through the ongoing Rudi Kundini, Pamoja Kundini (R01MH125746) study. Research from our team and others reveals that financial incentives can increase HIV testing, enhance linkage to care, and bolster treatment adherence. Yet despite this evidence, few financial incentive programs have been scaled for HIV and other urgent health threats sub-Saharan Africa. One reason for this ‘know-do’ gap is that simplified, automated, and secure software for implementing incentive programs at scale have yet to be developed, evaluated, and shared. To meet this need, through collaboration with the Tanzanian Ministry of Health and the Tanzanian technology firm Rasello, we developed an innovative, acceptable, reliable, and scalable software (the “Rasello mHealth system”) to automate incentive distribution. The software securely sends incentives through mobile money, monitors HIV care with biometric identification, sends SMS visit and laboratory test reminders, and integrates with partner databases via application programming interfaces (API). Extensive stakeholder testing has permitted iterative improvement of the software to seamlessly incorporate into clinic workflows. The software is being used in the parent R01 (R01MH125746), but it was not designed to align with software engineering best practices, FAIR4RS principles, or principles of open science, which will limit its impact. Given that this was unforeseen when R01MH125746 was awarded, the Rasello mHealth system is the focus of this application. We will first enhance the system’s source code, improve documentation and version management, and adopt build/test tools for community open-source development (Aim 1). We will then develop standard build and packaging tools to manage the Rasello mHealth system’s dependencies, produce containerized runtimes, and containerize the software for a tool registry (Aim 2). We will also refactor the mHealth system software to incorporate standard interfaces and data formats and replace custom code with standard, hardened libraries where indicated (Aim 3). Lastly, we will improve performance of the Rasello mHealth system through improved logging, monitoring, code profiling and optimization; leveraging parallelization and GPUs (Aim 4). The Rasello mHealth system is a valuable resource for the scientific community, combining many desired features in a user-friendly platform to securely scale financial incentives for a variety of priority health issues. As we enter Year 3 of R01MH125746, now is the ideal time to enhance the system according to software engineering best practices and principles of open science, a need which was unanticipated when the parent study was funded. These activities will improve the impact and sustainability of the software to successfully complete the parent project and has potential to benefit the broader scientific community. 1

PROJECT NARRATIVE Although financial incentives increase engagement in many beneficial health behaviors, the lack of software to scale delivery has hindered widespread adoption of this evidence-based approach. We will enhance existing research software designed to simplify, secure, and automate incentive delivery for HIV care to align with software engineering best practices and principles of open science, a need which was unforeseen when the parent study was funded. These activities will improve the impact and sustainability of the software to successfully complete the parent project and benefit the broader scientific community.