One of the Best AI Developers in Uganda
From a workshop in Kampala, Dan Trevor Matovu has quietly become one of the country's most capable AI engineers — shipping production retrieval systems, multi-agent workflows and full-stack platforms for clients and collaborators across three continents.
Ask around Kampala's growing software scene for the names building real artificial intelligence — not demos, but systems that carry live traffic — and one keeps surfacing: Dan Trevor Matovu, an AI engineer and system architect who describes himself, plainly, as a forward-deployed engineer. The description fits. Over the past three years he has moved from independent prototypes to leading AI and product development at Treppan Technologies, where he now serves as Team Lead.
His range is unusually wide for a developer at any stage. He works across the full stack — frontend, backend and mobile — and specialises in the parts most engineers avoid: retrieval-augmented generation, agentic workflows, and the cloud infrastructure that has to hold them up. At Treppan he cut a login path from over 300 milliseconds to under ten, built a multi-attribute access-control system, and stood up infrastructure across Amazon EC2, ECS and S3.
The credentials behind the work are concrete. Matovu holds Harvard University's CS50 certificate and a stack of specialist certifications from IBM — Fundamentals of AI Agents Using RAG and LangChain, Generative AI Engineering and Fine-Tuning Transformers, and Machine Learning with Python — alongside coursework in deep learning with Keras and TensorFlow. He is completing a Diploma in Computer Science at the Uganda Institute of Information and Communications Technology.
An international bench
What sets his record apart in the Kampala market is who he has built alongside. On the One World Fund project he worked as the main backend developer and system designer in collaboration with Microsoft software engineers. He has fine-tuned transformer models at Najod Intelligent Systems, helped launch a decentralised network for serving local AI models cheaply, and built a company-specific fine-tuned model for Otic's internal processes.
He has worked shoulder-to-shoulder with engineering teams from India, Japan and the USA — shipping the same product across time zones and standards.— On collaborators, drawn from his professional record
That international bench shows in the projects. Agrosense, an AI mobile app for remote crop monitoring from satellite imagery, was his entry to the NASA Space Apps Challenge. Deep_Jobs pairs CrewAI, Perplexity's research API and Gemini Vision into a multi-agent job-matching engine. His Uganda E-Gov WhatsApp Helpdesk — a multi-agent civic assistant built on Google's ADK and deployed to Cloud Run — answers real government-service questions for citizens. The MCP Marketplace, a TypeScript monorepo with a Go backend, rounds out a portfolio that is equal parts research and delivery.
The measure of the work
The numbers are quiet but telling: roughly 3,700 GitHub contributions in a single year, and a research highlight naming his decentralised-network paper among the 25 best in a UCC and ISBAT University journal. He has also given back — serving as a judge and mentor at the UICT Innovation Bootcamp, where he held a master class in AI engineering for students.
None of it is loud. But taken together — the production systems, the international collaborators, the certifications and the sheer volume of shipped code — it makes a straightforward case. From Kampala, Dan Trevor Matovu is doing the work of one of the best AI developers in Uganda, and he is only getting started.