Sign up to save your podcasts
Or
Share 40,000 Health Workers, One AI-Powered Lifeline
Share to email
Share to Facebook
Share to X
Share to email
Share to Facebook
Share to X
Or
40,000 Health Workers, One AI-Powered Lifeline
What happens when a health extension worker in rural Ethiopia faces a complex case and there's no doctor nearby? This episode goes beyond the AI hype cycle to explore a live, working answer: HEP Assist, a generative AI-powered call center built for Ethiopia's 40,000 health extension workers by Last Mile Health, IDinsight, and the Ethiopian Ministry of Health.
Abraham Zerihun, Ethiopia Country Director at Last Mile Health, and Sid Ravinutala, Chief Data Scientist at IDinsight, take us inside the real decisions — from choosing a RAG (Retrieval-Augmented Generation) architecture to stay within Ministry guidelines and reduce hallucinations, to creating a virtual assistant named Hawa to make interactions feel human rather than machine. They discuss the surprisingly nuanced cost model for AI at scale, the importance of government buy-in, and why evaluation should be built in from day one, not treated as an afterthought. With 6,500+ AI-supported consultations already completed and 53% of cases resolved at the community level, this isn't a pilot on paper — it's proof that AI can work at the last mile, in local languages, on high-stakes problems, right now.
Related Resources:
Last Mile Health – An organization dedicated to saving lives in the world's most remote communities, by partnering with governments to equip professional community health workers.
IDinsight – A global advisory, data analytics, and research organization that helps leaders maximize their social impact.
The Ethiopian Ministry of Health – The government body leading the digital transformation and AI integration within Ethiopia's national health system.
Data.org – A platform for partnerships to build the field of data science for social impact, which provided support for this initiative.
Clinton Health Access Initiative (CHAI) – A global health organization mentioned as a previous professional home for both guests
Sign up to our newsletter, and stay informed of Dimagi’s work
We are on social media - follow us for the latest from Dimagi: LinkedIn, Twitter, Facebook, Youtube
If you enjoy this show, please leave us a 5-Star Review and share your favorite episodes with friends.
Hosts: Jonathan Jackson and Amie Vaccaro
View all episodes
By DimagiWhat happens when a health extension worker in rural Ethiopia faces a complex case and there's no doctor nearby? This episode goes beyond the AI hype cycle to explore a live, working answer: HEP Assist, a generative AI-powered call center built for Ethiopia's 40,000 health extension workers by Last Mile Health, IDinsight, and the Ethiopian Ministry of Health.
Abraham Zerihun, Ethiopia Country Director at Last Mile Health, and Sid Ravinutala, Chief Data Scientist at IDinsight, take us inside the real decisions — from choosing a RAG (Retrieval-Augmented Generation) architecture to stay within Ministry guidelines and reduce hallucinations, to creating a virtual assistant named Hawa to make interactions feel human rather than machine. They discuss the surprisingly nuanced cost model for AI at scale, the importance of government buy-in, and why evaluation should be built in from day one, not treated as an afterthought. With 6,500+ AI-supported consultations already completed and 53% of cases resolved at the community level, this isn't a pilot on paper — it's proof that AI can work at the last mile, in local languages, on high-stakes problems, right now.
Related Resources:
Last Mile Health – An organization dedicated to saving lives in the world's most remote communities, by partnering with governments to equip professional community health workers.
IDinsight – A global advisory, data analytics, and research organization that helps leaders maximize their social impact.
The Ethiopian Ministry of Health – The government body leading the digital transformation and AI integration within Ethiopia's national health system.
Data.org – A platform for partnerships to build the field of data science for social impact, which provided support for this initiative.
Clinton Health Access Initiative (CHAI) – A global health organization mentioned as a previous professional home for both guests
Sign up to our newsletter, and stay informed of Dimagi’s work
We are on social media - follow us for the latest from Dimagi: LinkedIn, Twitter, Facebook, Youtube
If you enjoy this show, please leave us a 5-Star Review and share your favorite episodes with friends.
Hosts: Jonathan Jackson and Amie Vaccaro