Kenya's AI Pregnancy Chatbot Fields 15,000 Daily Questions

A text-based AI system in Kenya is handling 15,000 pregnancy-related questions daily, demonstrating how artificial intelligence can address critical healthcare gaps in resource-constrained settings.

The program, called PROMPTS, began accidentally when pregnant women started replying to automated appointment reminders from Jacaranda Health clinic networks. What started as fewer than 100 questions per day in 2020 has grown into a comprehensive maternal health service covering 24 Kenyan counties and reaching an estimated 700,000 users annually.

How the system works

PROMPTS operates entirely via SMS, requiring only a basic mobile phone—no smartphone, internet connection, or data plan needed. The AI analyzes incoming questions in English, Swahili, and Sheng (a Nairobi dialect mixing both languages), then prioritizes responses based on urgency.

About 7% of messages trigger immediate escalation. These bypass the AI entirely and route directly to one of 18 nurses monitoring the system around the clock. A nurse can review the patient's question history and call within minutes.

"If a mom in a rural part of Kenya starts bleeding at 2 a.m., she can't just Google it," says Jay Patel, Jacaranda Health's director of technology. "But this service runs completely over SMS. It's completely free to the user, and it's accessible to anyone who has a basic phone."

Why it matters

Kenya loses approximately 6,000 women annually to pregnancy and childbirth complications—16 deaths per day. Nearly one-third of these deaths stem from delays in seeking care. PROMPTS addresses this by helping women distinguish between normal pregnancy symptoms and medical emergencies, potentially preventing fatal delays. The system also reveals public health patterns: developers have identified correlations between extreme heat spikes and increased reports of headaches, swelling, and dehydration among pregnant users.

The human backstop

Unlike many AI healthcare applications drawing criticism for replacing human judgment, PROMPTS maintains significant human oversight. The 18-nurse team audits AI responses, monitors accuracy, and handles all urgent cases personally.

This human component raises costs—approximately $2 per user through pregnancy and the first postpartum year, absorbed by the nonprofit. But Javan Waita, Jacaranda Health's Kenya programs director, considers it essential: "The sooner a mother is able to access care, the sooner doctors will be able to understand the problem."

Concerns and limitations

Researchers note the system works best where functional healthcare infrastructure already exists. Javaid Iqbal Sofi, an AI policy researcher at Virginia Tech, raises a critical question: "What happens when the system gets a message wrong? Suppose a woman describes something urgent, but the system treats it as routine. Would someone catch that? How quickly?"

Smisha Agarwal, director of global digital health at Johns Hopkins University, sees PROMPTS as representing a fundamental shift in public health communication—from top-down messaging to patient-centered dialogue. "What the LLM really changes is the direction of conversation," she says. "Because now we can center it on what the mother actually wants to know, and in a private space on her own phone."

Jacaranda Health has expanded the system to parts of Ghana, Nigeria, and Tanzania, building language versions for Twi and Hausa speakers with localized nutritional and cultural context.

These details were first reported by NPR.