evidoria

← Back to browse

Good practice Imported

PROMPTS — Jacaranda Health's AI SMS Platform Guiding Millions of Kenyan Mothers Through Pregnancy and Postpartum Care

Kenya · Nairobi · See the Kenya profile · See the Nairobi profile

Evidence: Observational / pre–post Top 24% 73/100 · Ask Evidence Copilot about this practice

Jacaranda Health's PROMPTS service uses a Swahili-language AI model to answer 10,000-12,000 daily SMS questions from nearly 3 million enrolled Kenyan mothers, lifting completed prenatal-visit rates by 20% and routing 90% of flagged high-risk cases to timely hospital care.

~3 million
Women enrolled
23
Kenyan counties reached
10,000-12,000
Daily inquiries answered (per day)
~70 %
Inquiries resolved directly by AI
20 %
Increase in women completing 4+ antenatal visits
18 %
Increase in women completing 2+ postpartum visits
1.85 x
Higher engagement with postpartum family-planning services
90 %
AI-flagged high-risk cases reaching timely hospital care
0.74 USD
Cost per participant
PROMPTS — Jacaranda Health's AI SMS Platform Guiding Millions of Kenyan Mothers Through Pregnancy and Postpartum Care

Details

Maturity
Scaling
Promoter
Jacaranda Health
Period
2017–present
Keywords
maternal health, public health, AI messaging

Context

PROMPTS, run by Kenya-based nonprofit Jacaranda Health since 2017, is a Swahili-language AI messaging service that answers pregnant and postpartum women's questions by SMS and flags danger signs.

Objectives

The service aims to close the maternal-health information gap by giving women timely, accessible guidance through pregnancy and postpartum care, escalating urgent or complex cases to human clinical staff.

Activities

The AI resolves roughly 70% of 10,000-12,000 daily inquiries directly, with a team of about a dozen clinical nurses handling the rest; the service now reaches nearly 3 million enrolled women across 23 Kenyan counties, and more than 20 county governments use the linked PULSE dashboards for maternal-health planning.

Results

Funder-cited figures show a 20% increase in women completing four or more antenatal visits, an 18% increase in completing two or more postpartum visits, 1.85 times higher engagement with postpartum family-planning services, and 90% of AI-flagged high-risk cases going on to receive timely hospital care.

Conclusions

Jacaranda Health publishes responsible-AI principles covering data protection, human oversight of urgent cases and support for underserved languages; the AI's ~90-93% classification accuracy means a share of messages are still misrouted or need nurse correction, and the model is piloting in Ghana, Eswatini and Nigeria.

Implementation
Indicative cost
Medium (€50k–€500k)
Time to results
Long (> 3 years)
Staffing & skills
Jacaranda Health (nonprofit operator), ~12 clinical nurses handling escalated/complex cases, Partner county governments using PULSE dashboards

Conditions for success

  • Swahili-language AI model tuned to local maternal-health questions
  • Human nurse escalation pathway for complex or urgent cases
  • Published responsible-AI principles covering data protection, human oversight and underserved-language support
  • County-government adoption of linked planning dashboards (PULSE)

Common failure modes

  • AI classification accuracy of roughly 90-93% means a share of messages are misrouted or require nurse correction

Commonly funded by

Philanthropic / foundation funding National / regional programmes

Indicative funding routes for practices of this type — always check each programme's current calls and eligibility rules.

Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.

Data sources

Where this practice's information was retrieved from, and when.

Attachments

Similar practices you may find useful