^{By Rory Assandey and Dennis Maorwe}

Long before a consultation begins, people are already making decisions about their health.

It happens when a parent tries to understand a child's diagnosis for the first time. When a young woman searches for sexual health answers, she does not feel comfortable asking aloud. When a family weighs whether a journey to the nearest clinic is worth the cost, time and/or the uncertainty of what they will find.

Those moments are rarely captured in health records, yet they often determine whether care is sought in a timely manner or at all.

Across Africa, researchers are paying close attention to this space between everyday life and the health system infrastructure by asking how to reach people before they become patients. The starting point can be understanding why people remain beyond the reach of healthcare in the first place.

For Dr Dennis Maorwe of DPE Company, that question was deeply personal. He describes what unfolded after his father suffered a stroke. "As we sat in the hospital with forms we couldn't fill out and answers that came too slowly," he recalls, "I wondered: if I felt lost with all my training and networks, what must it be like for those without them?"

The experience informed his work with families affected by sickle cell disease in Kenya. Living with sickle cell disease involves much more than managing a medical condition. Families must navigate complicated treatment pathways, misinformation, stigma and limited access to trusted health information. In some communities, sickle cell disease is still misunderstood, leaving families to rely on advice that may misdirect, delay care and/or deepen fear.

Maorwe and his team addressed the challenges of these populations by developing a large language model (LLM) trained with input from patients, clinicians and specialists to deliver practical, culturally relevant health information. The goal was to make reliable information easier to access and easier to trust.

Technology works best when it strengthens existing relationships rather than replacing them, so community health workers became central to the project. They helped identify households who may not have appeared in clinic records. They followed up personally with families when text messages failed to reach them. They combined digital tools with the everyday realities of the communities they already knew well.

One mother said that receiving information and reminders in Swahili was the first time she truly understood her child's condition, describing the messages as feeling like "a friend checking in." Another caregiver reported that appointment reminders helped her keep up with clinic visits. They were simple interactions, yet they changed how families engaged with care.

Hundreds of kilometres away in Côte d'Ivoire, Rory Assandey, co-founder of La Ruche Health, and his team were discovering something similar, which led them to reconsider the design and approach of KIKO, an AI platform meant to help people find health information and navigate healthcare services.

Instead of asking health care questions directly, many people asked different kinds of questions. Some wanted to talk about anxiety, relationships, stress, sexuality and emotional wellbeing. Few described these concerns using clinical language. Most were simply looking for a place where they could ask questions openly without fear of judgement. Similar findings were emerging across the continent.

Over the course of the project, 1,653 people used KIKO, asking more than 14,000 questions through WhatsApp, Messenger, and the web platform. Those conversations led to 768 teleconsultations with healthcare professionals, a demonstration that digital conversations can become meaningful pathways to receiving clinical care.

Women accounted for 82% of all teleconsultation users. Women often describe midwives as trusted sources of advice, often turning to them before approaching other healthcare professionals with sensitive concerns. That insight reshaped the platform itself, connecting users directly with midwives, psychologists and sexual health specialists through services they were already comfortable using.

Mental health became the most requested teleconsultation service; moreover, psychology services registered a rebooking rate of 30%. A return for a second conversation suggests that people were finding support that they wanted to continue using.

The team reported that a young woman described KIKO as "my virtual friend", though she was well aware that she was interacting with artificial intelligence. Researchers concluded that trust could create a space where someone felt comfortable taking the first step towards seeking help.

The common threads among findings across these different settings and countries are that healthcare becomes more accessible when it reflects the way people actually live. Trusted information matters. Language matters. Familiar platforms matter. Relationships matter.

Artificial intelligence can strengthen each of these, but only when it is developed cooperatively with communities rather than simply delivered to them.

Through Grand Challenges Africa, an initiative of the SFA Foundation, researchers are exploring these observations alongside communities, health workers, and local institutions. The programme supports African-led innovation at an early stage, creating opportunities to test ideas, learn from the people they are serving, and generate the evidence needed for responsible and effective adoption.

About Rory Assandey

Rory Assandey is a computer engineer and co-founder of La Ruche Health in Côte d'Ivoire. He works at the intersection of artificial intelligence and healthcare, developing digital tools that help underserved communities access trusted health information and services. Through Kiko, an AI-powered health chatbot, he is helping make healthcare more accessible for young people.

Learn more about La Ruche Health through Online Doctor Consultations & AI Healthcare • La Ruche Health

About Dennis Maorwe

Dennis Maorwe is a physician and health innovator based in Kenya. His work focuses on using artificial intelligence to improve health outcomes and address public health challenges. Through his work on sickle cell disease, he is developing AI-driven tools that support disease management, empower patients with tailored information, and help improve quality of life.

Learn more about DPE Company through DPE | Health Engagement Infrastructure – From Outreach to Outcomes