+
+
+
+
Hi, I’m Elie M. Yene
+
+
+
+
/ CONTACT
Let's work together
If you're building AI systems that need to work reliably, not just once, let's talk.
If you're building AI systems that need to work reliably, not just once, let's talk.
Explore projects where I've delivered measurable business impact through scalable technical solutions.
Building backend systems in production
Green Revolution Initiatives (GRI)
Université Evangélique en Afrique (UEA)
Kikapu App
Digital Vision East Africa
Tools and languages I use to build systems
I spent three years building backend systems in production, then went back to school for an MSc in Applied AI at Warwick. I wanted to understand machine learning and LLM systems at the same depth I understood backend architecture: deep enough to make them reliable, not just functional. The backend experience carries over directly. I have built Spring Boot and NestJS microservices serving 5,000+ users, shipped authentication microservices with 300ms response times, and run production infrastructure on Keycloak, PostgreSQL, and Docker. My dissertation applies that same standard to AI. Does splitting a compliance agent into a multi-agent system actually make it more repeatable? I ran a pre-registered experiment across 41,400 scored runs and five open-weight model families to find out. The answer was model-dependent: repeatability is not something you get for free from architecture, and most of the inconsistency traced back to serving-stack cache state rather than randomness.
“Looking for AI/ML engineering roles in Canada where I can bring production backend experience and deeper knowledge of AI to build, evaluate, deploy, monitor and maintain reliable multi-agent and intelligent systems.”