Our Research Principles
At DPE, we believe that AI in health must be rooted in real-world needs. We research how AI can support safer, smarter, and more inclusive public health systems.
Equity Before Scale
We design first for underserved communities - community health workers (CHWs/CHVs), caregivers, and low-resource users - then scale. Inclusive by design, not as an afterthought.
Trust & Safety By Design
Every AI interaction earns trust through safety features, explainability, and user control - with governance that meets local regulations.
Open Methods, Not Black Boxes
We publish methods, share data responsibly, and build auditable, adaptable systems others can reuse - via open docs, datasets, and APIs where appropriate.
Ground-Truth, Co-Designed
We co-design with end users and validate in real settings - field data over theory - so tools work in Africa’s health systems.
Core Research Areas
Behavior Change Communication (SBCC)
Plain-language, culturally relevant messages across SMS/WhatsApp using Large Language Models, COM-B/SBCC and real-world A/B tests.
Trusted AI for Public Health
Safety guardrails, explainability, and interoperable services for auditable, governable deployments.
Community Feedback & Evaluation
We build citizen-signal pipelines and CHW/CHV feedback loops to monitor model quality, equity, and outcomes – and feed improvements back into the system.