I'm a fullstack engineer who works backwards from the architecture. I ship full-scale end to end applications and logic that sits between a database and a decision — APIs, pipelines, dashboards, and the occasional AI integration that has to actually behave in production.
Vaye is a four-app super-platform built for the Zimbabwean and South African markets. A passenger app, a driver app, an internal admin ops centre and a single Express + MongoDB backend cover ridesharing, deliveries, a peer-to-peer marketplace and an integrated wallet — with currency, payment provider and locale all driven by the user's region rather than hardcoded.
ClearedMind is a comprehensive mental wellness platform that pairs guided meditations, mood journaling, and clinically-shaped self-assessments with on-demand booking of vetted therapists. It ships from a single React/TypeScript codebase to web, iOS, and Android via Capacitor, on top of a custom native-abstraction layer that lets the same components feel native on a phone and graceful in a browser.
A production-ready AI agent service designed to plug into any application via adapters. It abstracts OpenAI, Anthropic Claude, and HuggingFace behind a single interface, supports tool/function calling against host apps, and ships with auth, rate limiting, full audit logging, and session management.
Most software I've shipped lives quietly inside a business — moving documents, tracking donors, scoring tenders, balancing assets. The interesting part was never the framework; it was figuring out the model that made the work *legible* to whoever had to use it the next morning.
I default to clean architecture, modular boundaries, and small interfaces. Not because it's fashionable — because it's what I can still understand at 2am when something is on fire.
I'm comfortable across the stack. C#/ASP.NET Core and Node/Express on one side, React/TypeScript on the other, MongoDB and SQL Server depending on what the data wants to be. AI integrations are part of the toolkit now, not a feature flag.