Back to Portfolio

Ethiopia Health System Deployment

Zenysis Technologies
Role Co-founder & Field Lead
Timeline 2015-2016
Focus Area Global Health, Data Infrastructure

Key Takeaway

Systems designed in Silicon Valley break in field conditions. Success requires continuous iteration with end users, adapting to unreliable connectivity, local workflows, and cultural context.

Problem Statement

Ethiopia's health system had fragmented data across multiple platforms—disease surveillance systems, health facility reporting, vaccination records, supply chain databases. Each system served a purpose, but none talked to each other. Decision-makers couldn't get a unified view of health outcomes, resource allocation, or emerging issues.

This fragmentation had real consequences:

The Ministry of Health needed integrated infrastructure that could unify disparate data sources, provide real-time visibility, and support data-driven decision making—all while working within the realities of limited connectivity, varying technical capacity, and existing workflows.

Technical Approach

As co-founder of Zenysis (YC W16), I led field deployment of an integrated health data platform in Ethiopia. The work required building technical infrastructure while navigating operational realities, ministry partnerships, and on-the-ground implementation challenges.

Platform Development

Field Deployment

Living in Addis Ababa for months, I worked alongside ministry teams to ensure the system worked in practice, not just in theory:

Impact

The platform enabled real-time health monitoring and resource allocation across Ethiopia:

Operational Improvements

Systemic Change

What I Learned

Field deployment is where theory meets reality. Systems that work perfectly in San Francisco break in unexpected ways in Ethiopia—from power outages to connectivity issues to browsers that haven't been updated in years. You can't design for field conditions from an office; you have to be there, observing failures, adapting continuously.

Users know their workflows better than you ever will. Ministry officials had deep expertise in health operations that our team lacked. When our design assumptions conflicted with their workflows, they were right and we were wrong. User-centered design isn't about surveys or personas—it's about continuous engagement with people actually doing the work.

Trust is earned through presence and responsiveness. Government partnerships don't succeed because of clever sales pitches. They succeed when stakeholders see you're committed for the long haul, responsive to problems, and genuinely invested in their success. Living in Addis, being available, showing up consistently—that built relationships that enabled deployment.

Technical infrastructure is only 30% of the work. The other 70% is training, change management, stakeholder alignment, process documentation, and ongoing support. Software doesn't transform systems; software plus sustained human effort transforms systems.

Sometimes slow is fast. We could have pushed for faster rollout, but taking time to ensure each deployment worked well, users were comfortable, and bugs were fixed created momentum. Rushed deployment would have generated resistance and failure. Patience created success.

Bridging Worlds

This project required constant translation between different domains:

Success came from operating in that gap—understanding technical possibilities and operational realities, translating between cultures and contexts, building systems that worked for actual users in actual conditions.

These lessons carried through to my subsequent work at Aclima (field deployment of environmental sensors), UN Global Pulse (translating data science for policy), Palantir (deploying AI in government), and Intramotev (autonomous systems in diverse rail environments). The core lesson remains: field deployment requires humility, iteration, and genuine partnership with end users.