Speaker: Dr.-Eng. Aurelle Tchagna Kouanou, University of Yaoundé I.

Title: When Data Is Scarce, Can AI Still Be Trusted? Building Agentic and Responsible AI for African Healthcare.

Abstract: Artificial intelligence is transforming healthcare, but its success often depends on large, high-quality, representative datasets, resources that remain limited in many African settings. This talk explores a critical question: when data are scarce, incomplete, heterogeneous, or uncertain, can AI still be trusted? We will examine the limitations of conventional predictive models in African healthcare, including data quality, bias, generalizability, uncertainty, and clinical trust. The presentation will introduce Agentic AI as a promising approach for building more responsible and adaptive healthcare systems. By combining specialized agents for data-quality assessment, prediction, uncertainty estimation, explainability, verification, and reporting, Agentic AI can move beyond simply producing a diagnosis or risk score. It can recognize insufficient information, communicate uncertainty, and support safer human decision-making. The talk will highlight practical research opportunities for developing trustworthy, context-aware, and responsible AI systems designed around the realities and needs of African healthcare environments while preserving the clinician’s central decision-making role.

Date & Time: [17th August 2026 | 8 PM (WAT)]

Venue: Online (Google Meet Link: meet.google.com/pjq-xson-qow )

• Rinel Foguen-Tchuendom, Bettermore Labs.