Senior technical lead. Cloud-native systems, end to end.
For the last five years I have been senior technical lead on Kubapay, a multi-tenant transit ticketing SaaS live in over 500 cities across France, Italy, South Africa and beyond. I lead a ten-engineer team out of India and work closely day to day with the product and engineering side of the team in Denmark.
I am also senior technical lead on Robogebra, a smart classroom platform for maths teachers, where I own the release cycle across web and mobile.
Alongside them I have delivered in legal tech, financial compliance (FATCA/CRS), industrial credentialing, logistics and edtech, and shipped AI systems of my own — MemoMind, a retrieval engine that answers from your own documents with a citation behind every claim, and UmarTransit-1B, a transit-domain language model published on Hugging Face.
Four kinds of problem account for most of the work clients bring me.
Cloud-native architecture — Java, Kotlin and Spring Boot microservices designed for scale and failure, deployed on AWS with Docker, Kubernetes and CI/CD that people other than me can operate. That includes multi-tenant platforms, where one system serves operators who each need it to behave as if it were built for them alone.
The clients on top: Angular and React on the web, Ionic and React Native on mobile, Electron on the desktop — one product across every screen, including the awkward parts like native integrations, store releases, and billing that reconciles across web, App Store and Play.
Modernization: AngularJS to Angular, old Java to current, monolith to Spring Boot microservices, done while the product stays live and paying customers keep using it.
And AI features that have to be right — retrieval grounded in your own documents with a citation behind every answer, and a fine-tuned model of your own when a general one is not accurate enough for your domain.