V
Loading...
Loading...
Loading...
Voice AI Transforming Community Health
A voice-enabled health assistant that helps community health workers triage patients in rural Tanzania, speaking natural Swahili and working completely offline.
75%
Faster Triage
60%
Fewer Unnecessary Referrals
10K+
CHWs Using System
500K+
Patients Assessed
Community health workers (CHWs) in rural Tanzania are the front line of healthcare for millions of people. But they often lack the training and tools to effectively triage patients and determine when referral to a clinic is needed.
Working with the Tanzania Ministry of Health and WHO Africa, we developed a voice-enabled health assistant that CHWs can use to guide patient assessments. The system speaks natural Swahili, works completely offline, and provides evidence-based recommendations.
The impact has been transformative: 75% faster triage times, 60% reduction in unnecessary referrals, and improved health outcomes for communities served.
Understanding the unique obstacles we're working to overcome.
Existing health tools are in English; CHWs and patients are more comfortable in Swahili.
Many CHWs have limited literacy, making text-based interfaces difficult to use effectively.
Rural areas where CHWs work have no reliable internet connectivity.
CHWs receive basic training but may not remember protocols for rare conditions.
The methods and techniques we've developed to address these challenges.
Natural language understanding and speech synthesis in Swahili for conversational interactions.
WHO-approved clinical decision support adapted for community health worker context.
Complete functionality without internet, with sync when connectivity available.
Extensive co-design with CHWs to ensure usability and trust.
Measurable outcomes from our research and deployments.
75%
Faster Triage
Average patient assessment time reduced from 20 minutes to 5 minutes.
60%
Fewer Unnecessary Referrals
More accurate triage means patients get appropriate care locally.
10K+
CHWs Using System
Over 10,000 community health workers across Tanzania using the assistant.
500K+
Patients Assessed
More than half a million patients assessed using the voice assistant.
Adeyemi, G., Hassan, F., et al.
Scaling the voice assistant to all 40,000 community health workers in Tanzania.
Partners:
Tanzania Ministry of Health, WHO Africa, USAID
Dr. Grace Adeyemi
NLP Lead
Fatima Hassan
Healthcare AI Lead
We're always looking for collaborators, partners, and talented researchers to advance this work.