● SOFTWARE ENGINEER / SYSTEMS DEVELOPER

KAMPALA / EAT

Engineering usefulintelligence.

I engineer diverse and specialized technological solutions, spanning embedded systems, AI/ML, data science, cloud-native backends, and mobile or web applications.

↓

WEB SYSTEMS ✳ AI & COMPUTER VISION ✳ MOBILE APPLICATIONS ✳MACHINE LEARNING ✳ DATA SCIENCE ✳ CLOUD DELIVERY ✳EMBEDDED SYSTEMS

Backend & Databases

FastAPI

Django

Flask

Laravel

Spring

Node.js

PostgreSQL

MySQL

MongoDB

SQLite

Web, Mobile & Cross-Platform

React

Next.js

Tailwind

Android

Flutter

Apple

Firebase

Cloud, Delivery & Tools

AWS

Docker

GitHub Actions

Git

GitHub

Autodesk Fusion

Arduino

Backend & Databases

FastAPI

Django

Flask

Laravel

Spring

Node.js

PostgreSQL

MySQL

MongoDB

SQLite

Web, Mobile & Cross-Platform

React

Next.js

Tailwind

Android

Flutter

Apple

Firebase

Cloud, Delivery & Tools

AWS

Docker

GitHub Actions

Git

GitHub

Autodesk Fusion

Arduino

FEATURED WORKS

AI + AGRICULTURE

CropWIT

A hybrid AI crop diagnosis platform that combines computer vision, field observations, weather, sensor data, and validated agronomic knowledge to produce evidence-backed diagnoses.

ROLETechnical Lead & System Architect

THE CHALLENGE

Turn a leaf image and field observations into a reliable crop diagnosis without treating visual similarity as the final answer or allowing a language model to invent agricultural conclusions.

THE APPROACH

Computer vision proposes candidate conditions. CropWIT then combines crop and variety, location weather, sensor readings, and validated Neo4j knowledge in a deterministic reasoning engine that ranks diagnoses and preserves the supporting, contradicting, and missing evidence.

Computer VisionFastAPIDjangoNext.jsNeo4jAndroidArduinoESP32

SELECT AN IMAGE TO EXPAND

AGENTIC AI + GUARDRAILS

Greenard

CropWIT's agricultural agent, to retrieve and explain validated diagnoses, answer follow-up questions, and guide farmers through the application.

ROLETechnical Lead & System Architect

THE CHALLENGE

Make detailed diagnosis evidence understandable to farmers without allowing a general-purpose language model to diagnose crops, alter rankings, or invent treatment guidance.

THE APPROACH

Built a bounded, tool-using agent with authenticated read-only retrieval, Pydantic-validated responses, controlled navigation actions, agricultural guardrails, and persistent PostgreSQL conversation memory.

OpenAI Agents SDKPythonPydanticPostgreSQLTool CallingAndroidGuardrails

SELECT AN IMAGE TO EXPAND

BACKEND + DELIVERY

Auth Microservice

A production-minded authentication service built by a five-person team, with CLI administrator bootstrap.

ROLETechnical Lead & System Architect · Five-person team

THE CHALLENGE

The team needed to build a dependable, plug-and-play authentication and role-based authorization microservice.

THE APPROACH

Created the base architecture, divided delivery through documented GitHub issues and milestones, protected the main workflow, reviewed and integrated pull requests, tested the combined service, and published it through CI/CD.

FlaskPostgreSQLDockerGitHub ActionsRailway CLI

Explore repository ↗

SELECT IMAGE TO EXPAND

ANDROID APPLICATION

MenuLive

A food-discovery Android application built with Jetpack Compose, combining local and cloud persistence with testable components.

ROLETechnical Lead & System Architect · Five-person team

THE CHALLENGE

The team needed to make discovery immediate, simple, and only a click away while coordinating customer and vendor locations, cloud synchronization of orders, proper dependency injection, and a modern native interface.

THE APPROACH

Established the Compose architecture and starter project, split the work into owned deliverables, reviewed and requested revisions on pull requests, integrated contributions on the development branch, and tested the release before publishing.

KotlinJetpack ComposeHiltRoomFirebaseFirestore

Explore repository ↗

SELECT AN IMAGE TO EXPAND

SUPPLY CHAIN MANAGEMENT SYSTEM

Clothe SCM

A supply-chain platform that connected raw materials, production activity, performance signals, and final customer transactions.

ROLETechnical Lead & System Architect · Five-person team

THE CHALLENGE

The team needed to turn a multi-stage operational process (raw-material sourcing, supply and delivery, manufacturing and production, inventory, and sales) into one coherent system with measurable performance and integrated payments across the chain.

THE APPROACH

Defined the Laravel foundation and system architecture, coordinated deliverables through GitHub and ClickUp, reviewed and integrated team contributions, and tested the complete workflow with Flutterwave card and Mobile Money payments.

LaravelBootstrapBladeTailwindKPIsMachine LearningSpring BootFlutterwave

Explore repository ↗

SELECT AN IMAGE TO EXPAND

ENTERPRISE SYSTEMS EXPERIENCE

Enterprise Resource Planning& Management Systems

Experience on both sides of ERP adoption: trained users to work confidently with business systems and extended and customized open-source platforms around real operational needs.

ERPNext ERPNext

Made business operations easier to manage.

Built custom modules with the Frappe Framework to help small enterprises replace manual processes with accurate, accessible workflows and clearer financial decision-making.

FrappePythonJavaScriptDocTypesServer ScriptsUser Training

Odoo

Extended open source around businesses.

Customized, extended, and implemented Odoo through purpose-built modules and complex integrations, with attention to maintainable architecture, adoption, and scalable operations.

PythonXMLJavaScriptCustom ModulesIntegrationsUser Training

SYSTEM DOCUMENTATION AND REVERSE ENGINEERING

ECOMMERCE + SYSTEM ARCHITECTURE

Eferno

A high-level software design documentation for a scalable, trust-aware ecommerce platform, translating user needs into a structured microservices and cloud-native system architecture.

RESEARCH

Collected and analyzed user feedback through interviews, surveys, and questionnaires to identify ecommerce pain points, expectations, and trust requirements.

DOCUMENTATION

Defined the system overview, scope, service domains, communication patterns, layered architecture, cloud infrastructure, security, observability, and deployment model.

InterviewsSurveysQuestionnairesRequirementsMicroservicesCloud Architecture

View restricted documentation ↗

Access is restricted. Request access from me to view the document.

SELECT AN ARCHITECTURE IMAGE TO EXPAND

PUBLISHED DEVELOPER DOCUMENTATION

CROPWIT KNOWLEDGE GRAPH PIPELINE

Mintlify

Generated an OpenAPI 3.1 contract from Django API definitions and Pydantic models for 14 paths and 14 operations, then reused it in a protected Mintlify portal so the API reference stays aligned with the implementation.

MDX and Mermaid map components, data flows, and release boundaries.They separate versioned knowledge from runtime diagnosis evidence.

Open protected documentation ↗

OpenAPI

Swagger

GitBook

ACCESS HUB AUTHENTICATION SERVICE

Centralized 30 API operations across 21 paths in one OpenAPI specification, rendered through Swagger UI for live Flask testing and published with GitBook for engineers, frontend developers, stakeholders, and non-technical leads.

Interactive testing also exposed and traced a live application-model and PostgreSQL-schema mismatch to unapplied migrations.

Open public documentation ↗

ENGINEERING RANGE

WEB

Web systems

React, Next.js and Product UI/UX.

Applied intelligence

Machine learning, computer vision, llms, agents and knowledge graphs.

APP

Mobile applications

Android, Jetpack Compose, Flutter, iOS and Firebase.

API

Backend platforms

FastAPI, Django, Flask, Laravel, Spring and Node.js.

Data foundations

PostgreSQL, MySQL, MongoDB, SQLite and Neo4j.

PCB

Embedded systems

Arduino, Autodesk Fusion, sensors, circuits and field hardware.

Python

Kotlin

Java

PHP

TypeScript

JavaScript

C++

ENGINEERING PHILOSOPHY

I don’t just connect technologies. I connect decisions, people, and outcomes.

I am a software engineer and innovator passionate about leveraging technology to solve real-world challenges in Africa’s digital race and the global AI revolution.

I believe that the intersection of Artificial Intelligence and robust data management is the key to unlocking the potential of Ugandan enterprises.

THE ENGINEERING LOG

More signals from
the workshop.

Experiments, supporting systems, and focused builds beyond the featured case studies.

01Business Validation ServerSpring Boot · Maven · layered architecture↗02CropWIT Knowledge GraphDjango · MongoDB · JSONL · Neo4j / Cypher↗03AI Agent ExperimentsCustom tools · planning · reasoning↗04Odoo CustomizationPython / XML · PostgreSQL · 2FA · SMTP↗05TriconaNews · events · static web · GitHub Pages↗